About 36 min onwards is great listening
BRN Discussion Ongoing
- Thread starter TechGirl
- Start date
Taproot
Regular
Am I missing something here ?Just thinking out loud to myself ,
So we know we have some fancy radar offering ,
View attachment 96509
And we know that ASICLAND is designing an AKD2500 ASIC for tapeout to provide validation samples ,
View attachment 96511
And we know that Brainchip have gone to great lengths to tell us that it does not include customer orders ,
View attachment 96512
And we know about our partnership with Forward Edge and their little circle of defence and military friends ,
View attachment 96513
Including L3 Harris ,
View attachment 96514
Now , what we probably don’t know is L3 Harris have the below radar system ,
View attachment 96515
And looking at this report , which is apparently fact checked ,
View attachment 96516
And within the report we see that Brainchip and L3 Harris have a strategic partnership to intergrate Akida 2 into the Silent Knight radar system and the prototype is scheduled for Q3 2026 . I also believe that Brainchip will receive a small portion of the $18.5M contract , so it might pay to keep our eyes open to see how this one plays out .
View attachment 96517
Neuromorphic Computing for Defense Market Research Report 2034
The neuromorphic computing for defense market was valued at $1.8 billion in 2025 and is projected to reach $9.7 billion by 2034, growing at 20.5% CAGR.marketintelo.com
When did Brainchip announce a strategic partnership with L3Harris ?
"February 2026
BrainChip Holdings Ltd. - Announced a strategic partnership with L3Harris Technologies to integrate the Akida 2 neuromorphic processor into L3Harris's AN/APQ-187 Silent Knight radar system for autonomous target classification, with initial prototype delivery scheduled for Q3 2026 under a U.S. Army SBIR Phase III contract valued at $18.5 million. "
I listened to the Pitt Street Research Semiconductor Conference 2026 with BrainChip this morning, here's my transcript from my iPhone (not perfect... but you get the drift). I believe it will be coming out officially over the next couple of days (from Pitt Street) I'll also post the slides shortly.
*********-------------*********
Here we are. Hello, Mark. Hey, Sean. Um, so last presentation for today. So we're very happy to hear from you, Sean. Please welcome Sean Hehir, CEO.
Thank you, Mark. Thank you, everybody, for your attention at the end of the day. First of all, let me offer my apologies that I could make it there in person. I was, but I've had an injury that keeps me off airplanes, but I will be in Australia in about 5 weeks for our shareholders there.
Um, I'm a little bit different. I joined the last couple of presentations where I've talked a little bit about the industry, knowing this is a semiconductor conference. So maybe before I go in, let me just take a moment and position the company and glad to take more questions as we go through it about the company. For size wise, we're roughly about 100 professionals who will exit this year roughly about 100 professionals, mostly scientists and engineers that develop the world's leading technology around Edge AI. And specifically, we’re neuromorphic in nature, which is extremely low power and very high performance. Most of our workforce is, um, is scientists and engineers. You know, I would say 90% of the organisation is like that. Uh, we're listed in Australia, as Mark said, uh, our technical workforce is mostly in California where I'm based right now, and India and France, but of course, we have smattering of sales personnel around the world.
We make, at the simplest level, make intellectual property for the technology people in the audience that when people build computer chips, whether it's an ASIC or an SOC, we sell IP blocks. We've got 2 generations in the market, and we have a 3rd one coming out this year. And late last year, we announced our expansion into offering chips themselves. Our 1st production chip is available. We've already sold several 1000s and 10s of 1000s of those, and we have, um, there will be available this summer for others and we're taking orders on those right now.
And what you can expect from this company going forward for the next 5 to 7 years is continued expansion on our IP and chips as well as reference models and, um, and reference architectures as well. And with that, I'd like to go to the 1st slide.
So a little bit about, you know, the, uh, the industry with, let's just talk about that, just driving a lot of the activity around the market conditions for a company like us.
And the AI chip boom, and if you pay attention to this industry, specialised silicon is something that's really happening more and more in this industry, creating market opportunities for companies like us to sell either IP or chips. You can run AI models and for those who are maybe not AI, um, uh, aware, a model or an algorithm is similar to think of like a program. You can run AI models and basically any silicon. You can run on a CPU. It just would be very power intensive and probably pretty slow. You can run them on GPUs. But, really, the right thing to do is get specialised silicon. And so companies like BrainChip that we create with, they call an NPU neural processing unit, runs models exceptionally well. And what we have, obviously, is a piece of technology that uh, is scalable, portable meaning you can take it to any fab around the world and you can make it your chips larger or smaller. So it's very portable to make these specialised chips that's powering all the use cases that I'll describe later on. Secondly, I was on for when I saw one of my, um, good friends on there, Kobe earlier, I joined for a little bit in the memory bottleneck. Our key to our, and somebody asked that question about that. about memory and they're working with BrainChip. There's something about our architecture where we put the memory right next to the compute element. So we really don't have memory bottlenecks by the way we have this architecture structured up, which is a more and more important trend in this space. The energy crisis, I don't think I need to talk very much about it, not only in the data centre, but the edge, and as used cases are getting more and more proliferated around the world, and people are demanding more and more functionality - the ability to get the most business value, the most functionality out of low power, and that's where we shine. And then, of course, latency, you know, when you talk about edge, and we'll talk a little about some use cases now, the ability to do inference outside of that, of the data centre, and avoid that latency and the security issues where privacy make a big difference. So efficiency is basically the new currency and shaping edge AI right now, and that's really where we sit.
The next slide, please. So I'm going to hit some trends in where we we, we address them at a simplest level and maybe you build out the next slide if you don't mind, then I'll do both at the same time to make sure we keep on track. There's this thing called physical AI that's on the industry, which is doing inference in right near at the activity next to a sensor or sensor fusion and things like that. And of course, when you make edge technology like us, BrainChip, that's exactly what we do, which is physical AI enabling that possibility.
Other big trends in AI is moving LLMs to the edge. We have an offering coming out later this year when you've got generative AI models that we're building an underlining impute to support them exceptionally well. We were leaning into a state, a set of models called state-based models. We'll support transformers as well. But the combination of state-based models and our underlining MPU will have world-class performance at a very small form factor to enable a lot of incredible use cases. Of course, the big drive around data privacy and sovereignty, and you hear these things about ethical AI and things like that, the ability to do our inference locally, and of course, that's exactly what we do. And probably the bigger one is what they call on-device inference. You know, if you paid attention to the industry at all in 25, there were some major trends that happened.
Of course, NVIDIA made a big deal with Gronk and paid $20 billion and acknowledging that you don't need a GP to do inference. That was like the big a-ha. Of course, we've known that for many years, but the industry's waking up. And then once you get over to that, there's a whole awakening in the industry that not only do you not need a GPU to do inference, you can do it locally and you should do it locally. And anybody in the tech world knows that everything starts centralised and it decentralises and you get to that fine balance. And that's where we are in this industry right now.
Next slide, please. So at its simplest level, and I'm going to just, I've only got a handful of slides and we'll open up some questions, which I see are coming in the chat window. When people ask a lot of things that are happening, where are we seeing most of the demand for this company? And we could talk about general edge markets and things like that. And I'm going to show some examples in there.
We're seeing tremendous demand and we're working. I can give some examples in the Q and A section about in the defence industry. And I know there were some discussions earlier about drones and others. There's a lot of things going on in the defence industry, create this opportunity right now, and I'll save them for a subsequent slide. Obviously, in wearables too. If you're going to do AI on a wearable device, you've got to hit all the points I had in the previous slide, which is privacy, latency, low power, because if you're going to have a wearable device, you don't want to be charging it every day. You want to charge it once a week. And when you're a neuromorphic or event-based like us, you get that incredible battery power that's allowed or to allow you not have to charge them, which is ideal for wearables. And then, of course, what I mentioned earlier is voice activated AI or LMs on the edge. We have an offering that addresses all of that, puts us right in the sweet part for these markets that are growing and growing quickly.
Next slide, please. So I'm going to just give you a couple quick examples and then maybe take some of the questions that are showing up in the chat window for us around. So one is a very specific example of what we're seeing in the strong demand in around here, which is around mobile radar. And we have a reference architecture in here. And in this case, the use case that developed from us was some work that we do with Raytheon and the US Air Force Research Lab. Where the challenge was, most traditional radar is very old - decades old. It's not very portable, and today's kind of defence environment, you need portability, and you need autonomy. And most importantly, you need a relatively low cost. And that's none of those things existed in traditional radar. And quite frankly, they're looking for additional functionality. Not only is something coming, they would like to be really good classified. We developed this technology and, you know, which is part of our underlying MPU and some of the algorithms associated with that. And we're working with those leaders I just mentioned in several other tier one defence contractors. at this very moment about proliferating that technology into markets.
Next slides. I mentioned a little bit earlier about the ability to have a private voice companion to do something.
Imagine the possibilities when you could talk to things like appliances, like medical devices, with a full-fledged LLM, think about your kind of a mini Claude or a mini chat GPT untethered to the cloud. You're not going to be able to ask at everything, but anything in the relevant conversation, you can do it. What does that mean?
Lower internet costs, much quicker response business. Um, the manufacturers really like that because it lowers their support costs. All these are realities and we're working with several companies right now about the possibilities of putting these kind of LLMs into their devices.
The next slide please? I want to bring out this one around the wearables market, to try to say that at the simplest level is a really good example for where we are to bring this to life, which is the possibility about a wearable. Now with Onsor, which is a company out of Oman, they've created these seizure, excuse me, epileptic seizure, prediction glasses. And in there, they have an Akida 1500, which is our chip, and they have a sensor in there, and it allows them with over the high 90%ile accuracy to predict if somebody's about to have a seizure within one hour. And I went to their kickoff event of roughly a year ago, and I heard a lot of stories from a lot of people during that time who were afraid to drive, afraid to leave their home, but now can do that because of these glasses. Their clinical trials are going very, very well, and Oman, their next plan is to roll that out to the rest of the region and then, of course, the rest of the world. You can see the tremendous impact to do this, and it's only possible if you have a chip that could do the inference locally, securely so they don't have their data go floating around, and an extended battery life. Right now, they're only charging these glasses once a week. Their previous trials with other vendors was every 12 hours. So you can see the impact of our kind of technology there.
In the next slide, if I could, please, another example about where we're going into the market and seeing success, which is another demanding environment, which is space. We have a licensee, I mentioned earlier, that we have a license model. We also have a model where we sell chips. FrontGrade, which is a company that is, um, that creates chips specifically for, um, for space agencies and large aircraft companies in Europe is a licensee of ours, and they're creating these chips that are space grade, rad hardened for those who are familiar with that term. that allow you to do all kinds of AI function AI functionality, autonomous in space, some kind of object detection, perhaps some kind of logistical support, navigation support. Um, that chip will build later this year and start to roll it in mass production of the licensee we work with now for a couple of years.
And these are just some of the examples of the possibilities that you can do with our technology and to use cases that were, quite frankly, not even possible a few years back.
In the next slide , please. Oops. And with that, I'm just gonna kind of bring it to a close and maybe talk a little bit about the company, right? In the end, we are a leader in edge AI technology, again, that provides only the IP, the silicon, the software and solutions. And I'll come back to that.
The market trends are undeniable. The pace of Edge AI, and I think I heard some of the comments from the previous speakers. And if anybody wants to talk about that, we can.
And we're getting strategic traction with companies in these verticals and others right now. Now, before I stop talking about this, and maybe take a few questions and try to keep us back on our time allocation. is that I want to talk a little bit about the left because it's easy to say provide a chip or even provide IP. But one of the things that's really important when you go through a company like us and get the breakthrough performance, you want to make it very easy for people to adopt your technology. So we spend a fair amount of capital working on our software stack in our model zoo to allow people to either bring their models on easily onto our platform, or adopt our models into their own. You can develop your models using industry standard frameworks like tensor flow or pie torch or onyx for those who are AI savvy and put it on. You don't have to learn anything at new to do that because if you've got a bunch of engineers, it's very easy to adopt our technology.
So in the end, we're looking for that breakthrough performance and ease adoption. And with that, I'd like to close my prepared remarks and take some questions, please. Um, yeah, I'll kick it off, Sean, with one question that was emailed in yesterday.
Question - There was some news out of IBM where they paired Akida chips with some IBM products and also some products from Palantir to solve a real world problem for the U.S. Army. So those are really exciting sort of applications. But what does it tell you about the maturity for, for, you know, Akida 1st form, obviously, but also neuromorphic processing, and and about the, um, the verticals, the, the best verticals for BrainChip to generate revenues from.
Sean Answer - Yeah, listen, it's a wonderful use case. It tells me a lot to answer your question about the maturity of it, which is it's very mature and ready for prime time. That's what we're seeing right now. And I could literally use my whole 20 minute allocation mark to talk about that because one thing I would lean into with that, which is so interesting. Most people tend to think of edge AI, even my comments earlier was more discrete, if you will, from the data centre. But with that example, it's a wonderful example where the right tool for the right job. If you read things that were publicly announced, the raw data was kind of overloading and squashing the Palantir system. By putting a Akida to next to those sensors and really turning all this raw data into useful information that was then fed into, it not only allowed it to do it, it allowed it to do it more efficiently and quicker.
And so it's that kind of working relationship instead of just discreetly in some of the examples I had. But this example of working, um, collaboratively with a data centre or centralised services is really where we think this thing is going to have a 2nd set of capabilities in there. And that is a wonderful example.
Now, also specifically about, about, you know, the vertical comment that you made Mark in there too. I'm going to call it a couple more. I mentioned earlier today, just a few moments ago, that we had an engagement with Raytheon that's going incredibly well. Again, a very demanding environment and customer. Um, you know, so there's two.
We also, when we announced the 1500 ship last fall, we quickly announced with a company called Parsons, which is a tier one defence contractor in the in the US that placed an order for several 1000 chips in a multi-year relationship that's going very well, and that will grow in more, many more 1000 chips will call. Another tier one that's putting them into real products and pushing them out to field.
And as recent as last week, there's a company called ForwardEdge ASIC that announced that they have chosen us, not for a particular chip at this point, but for future chips and ForwardEdge ASIC, if you pay much attention to them, is a wholly owned security of Lockheed, a very large tier one defence contractor.
To answer your question about the maturity and the readiness is for real applications, all of those are very demanding companies and we're tier one, so the technology is ready because they're putting it in it.
Thank you, Sean. Some questions online. Uh, yeah, correct. 2 questions online.
Question 1st question for you, Sean, is how would you see his commercial timeline for significant revenue increase?
Sean Answer - Yeah, we believe this year will be significantly more than anything you've seen in the past. And, uh, and we're going to see 26, 27, 28 is the years that we really take off. Why do I believe that?
One is obviously some of the edge market stuff that I had to go very quickly, but I'll take just a moment here, again, cognisant of the time. The use cases, if anybody went to CES, I use that as an example is the best. The number of edge use cases at CES, I would say, were 10 times a year before, and the year before was 10 times.
Not 10%, 10 times. So real products, real things are coming out, real demands.
Um, I've been with this company about 4 years. When I 1st joined most of the conversations were, hey, this is interesting, help me understand it. Maybe I can figure out what to do. Now the conversations I have with customers or like, here's our product plans, help us understand if you can help us do it. So everybody's got the mindset of implementation. So I believe those are the things.
Secondly, what I believe 26, 27 or 28 are our years is the broadening strength of our portfolio. When I said I came here, we had one generation of IP. We have two. We'll exit this year at three. We have chips. Now we have modules now. All the things are some of the previous speakers talked about that people go out and try things and deploy before they make IP decisions. as well as reference designs that we're putting out. You put those 2 factors together, a much stronger product set, a much stronger company, and stronger market forces, we believe this is the year 26, 27, 28 to watch us.
One more question - What part of friendship roadmap are customers responding to most strongly right now? Is it the current capability offerings or what's coming next from BrainChip?
Sean Answer - It's both, and I don't want to kind of give a political answer, but here's here's what happens in technology. I've been in technology for many decades. And this is where you, some of the failures that happen in this industry for, you know, companies or competitors of us or similar spaces. This is the ultimate business to business sale. It is not business to consumers. You're selling to people that you're enabling The customers of BrainChip were enabling their products, which is they're going to sell for revenue, right? So it's a very strategic sale. And they don't buy it on a single point. Yes, it has to solve their problem today with that single point. But they want to see a vision and a strategy 5 and 7 years out. And we provide that. So, yes, their technology today that will solve their problems and that's why we're seeing attraction, but they want to see if we have the vision 5 or 7 years that aligns with them. And so that's the answer to it. We solve it today, and we're also have a roadmap that is compelling for them to say, we can go with them and grow with BrainChip for the subsequent years.
All right, thank you very much, Sean. Do we have any questions in the room here?
If not, then, well, wrap it up, Sean. Thank you very much for your time. Thank you, Mark. and I look forward to seeing you there soon. Thank you very much.
*********-------------*********
Here we are. Hello, Mark. Hey, Sean. Um, so last presentation for today. So we're very happy to hear from you, Sean. Please welcome Sean Hehir, CEO.
Thank you, Mark. Thank you, everybody, for your attention at the end of the day. First of all, let me offer my apologies that I could make it there in person. I was, but I've had an injury that keeps me off airplanes, but I will be in Australia in about 5 weeks for our shareholders there.
Um, I'm a little bit different. I joined the last couple of presentations where I've talked a little bit about the industry, knowing this is a semiconductor conference. So maybe before I go in, let me just take a moment and position the company and glad to take more questions as we go through it about the company. For size wise, we're roughly about 100 professionals who will exit this year roughly about 100 professionals, mostly scientists and engineers that develop the world's leading technology around Edge AI. And specifically, we’re neuromorphic in nature, which is extremely low power and very high performance. Most of our workforce is, um, is scientists and engineers. You know, I would say 90% of the organisation is like that. Uh, we're listed in Australia, as Mark said, uh, our technical workforce is mostly in California where I'm based right now, and India and France, but of course, we have smattering of sales personnel around the world.
We make, at the simplest level, make intellectual property for the technology people in the audience that when people build computer chips, whether it's an ASIC or an SOC, we sell IP blocks. We've got 2 generations in the market, and we have a 3rd one coming out this year. And late last year, we announced our expansion into offering chips themselves. Our 1st production chip is available. We've already sold several 1000s and 10s of 1000s of those, and we have, um, there will be available this summer for others and we're taking orders on those right now.
And what you can expect from this company going forward for the next 5 to 7 years is continued expansion on our IP and chips as well as reference models and, um, and reference architectures as well. And with that, I'd like to go to the 1st slide.
So a little bit about, you know, the, uh, the industry with, let's just talk about that, just driving a lot of the activity around the market conditions for a company like us.
And the AI chip boom, and if you pay attention to this industry, specialised silicon is something that's really happening more and more in this industry, creating market opportunities for companies like us to sell either IP or chips. You can run AI models and for those who are maybe not AI, um, uh, aware, a model or an algorithm is similar to think of like a program. You can run AI models and basically any silicon. You can run on a CPU. It just would be very power intensive and probably pretty slow. You can run them on GPUs. But, really, the right thing to do is get specialised silicon. And so companies like BrainChip that we create with, they call an NPU neural processing unit, runs models exceptionally well. And what we have, obviously, is a piece of technology that uh, is scalable, portable meaning you can take it to any fab around the world and you can make it your chips larger or smaller. So it's very portable to make these specialised chips that's powering all the use cases that I'll describe later on. Secondly, I was on for when I saw one of my, um, good friends on there, Kobe earlier, I joined for a little bit in the memory bottleneck. Our key to our, and somebody asked that question about that. about memory and they're working with BrainChip. There's something about our architecture where we put the memory right next to the compute element. So we really don't have memory bottlenecks by the way we have this architecture structured up, which is a more and more important trend in this space. The energy crisis, I don't think I need to talk very much about it, not only in the data centre, but the edge, and as used cases are getting more and more proliferated around the world, and people are demanding more and more functionality - the ability to get the most business value, the most functionality out of low power, and that's where we shine. And then, of course, latency, you know, when you talk about edge, and we'll talk a little about some use cases now, the ability to do inference outside of that, of the data centre, and avoid that latency and the security issues where privacy make a big difference. So efficiency is basically the new currency and shaping edge AI right now, and that's really where we sit.
The next slide, please. So I'm going to hit some trends in where we we, we address them at a simplest level and maybe you build out the next slide if you don't mind, then I'll do both at the same time to make sure we keep on track. There's this thing called physical AI that's on the industry, which is doing inference in right near at the activity next to a sensor or sensor fusion and things like that. And of course, when you make edge technology like us, BrainChip, that's exactly what we do, which is physical AI enabling that possibility.
Other big trends in AI is moving LLMs to the edge. We have an offering coming out later this year when you've got generative AI models that we're building an underlining impute to support them exceptionally well. We were leaning into a state, a set of models called state-based models. We'll support transformers as well. But the combination of state-based models and our underlining MPU will have world-class performance at a very small form factor to enable a lot of incredible use cases. Of course, the big drive around data privacy and sovereignty, and you hear these things about ethical AI and things like that, the ability to do our inference locally, and of course, that's exactly what we do. And probably the bigger one is what they call on-device inference. You know, if you paid attention to the industry at all in 25, there were some major trends that happened.
Of course, NVIDIA made a big deal with Gronk and paid $20 billion and acknowledging that you don't need a GP to do inference. That was like the big a-ha. Of course, we've known that for many years, but the industry's waking up. And then once you get over to that, there's a whole awakening in the industry that not only do you not need a GPU to do inference, you can do it locally and you should do it locally. And anybody in the tech world knows that everything starts centralised and it decentralises and you get to that fine balance. And that's where we are in this industry right now.
Next slide, please. So at its simplest level, and I'm going to just, I've only got a handful of slides and we'll open up some questions, which I see are coming in the chat window. When people ask a lot of things that are happening, where are we seeing most of the demand for this company? And we could talk about general edge markets and things like that. And I'm going to show some examples in there.
We're seeing tremendous demand and we're working. I can give some examples in the Q and A section about in the defence industry. And I know there were some discussions earlier about drones and others. There's a lot of things going on in the defence industry, create this opportunity right now, and I'll save them for a subsequent slide. Obviously, in wearables too. If you're going to do AI on a wearable device, you've got to hit all the points I had in the previous slide, which is privacy, latency, low power, because if you're going to have a wearable device, you don't want to be charging it every day. You want to charge it once a week. And when you're a neuromorphic or event-based like us, you get that incredible battery power that's allowed or to allow you not have to charge them, which is ideal for wearables. And then, of course, what I mentioned earlier is voice activated AI or LMs on the edge. We have an offering that addresses all of that, puts us right in the sweet part for these markets that are growing and growing quickly.
Next slide, please. So I'm going to just give you a couple quick examples and then maybe take some of the questions that are showing up in the chat window for us around. So one is a very specific example of what we're seeing in the strong demand in around here, which is around mobile radar. And we have a reference architecture in here. And in this case, the use case that developed from us was some work that we do with Raytheon and the US Air Force Research Lab. Where the challenge was, most traditional radar is very old - decades old. It's not very portable, and today's kind of defence environment, you need portability, and you need autonomy. And most importantly, you need a relatively low cost. And that's none of those things existed in traditional radar. And quite frankly, they're looking for additional functionality. Not only is something coming, they would like to be really good classified. We developed this technology and, you know, which is part of our underlying MPU and some of the algorithms associated with that. And we're working with those leaders I just mentioned in several other tier one defence contractors. at this very moment about proliferating that technology into markets.
Next slides. I mentioned a little bit earlier about the ability to have a private voice companion to do something.
Imagine the possibilities when you could talk to things like appliances, like medical devices, with a full-fledged LLM, think about your kind of a mini Claude or a mini chat GPT untethered to the cloud. You're not going to be able to ask at everything, but anything in the relevant conversation, you can do it. What does that mean?
Lower internet costs, much quicker response business. Um, the manufacturers really like that because it lowers their support costs. All these are realities and we're working with several companies right now about the possibilities of putting these kind of LLMs into their devices.
The next slide please? I want to bring out this one around the wearables market, to try to say that at the simplest level is a really good example for where we are to bring this to life, which is the possibility about a wearable. Now with Onsor, which is a company out of Oman, they've created these seizure, excuse me, epileptic seizure, prediction glasses. And in there, they have an Akida 1500, which is our chip, and they have a sensor in there, and it allows them with over the high 90%ile accuracy to predict if somebody's about to have a seizure within one hour. And I went to their kickoff event of roughly a year ago, and I heard a lot of stories from a lot of people during that time who were afraid to drive, afraid to leave their home, but now can do that because of these glasses. Their clinical trials are going very, very well, and Oman, their next plan is to roll that out to the rest of the region and then, of course, the rest of the world. You can see the tremendous impact to do this, and it's only possible if you have a chip that could do the inference locally, securely so they don't have their data go floating around, and an extended battery life. Right now, they're only charging these glasses once a week. Their previous trials with other vendors was every 12 hours. So you can see the impact of our kind of technology there.
In the next slide, if I could, please, another example about where we're going into the market and seeing success, which is another demanding environment, which is space. We have a licensee, I mentioned earlier, that we have a license model. We also have a model where we sell chips. FrontGrade, which is a company that is, um, that creates chips specifically for, um, for space agencies and large aircraft companies in Europe is a licensee of ours, and they're creating these chips that are space grade, rad hardened for those who are familiar with that term. that allow you to do all kinds of AI function AI functionality, autonomous in space, some kind of object detection, perhaps some kind of logistical support, navigation support. Um, that chip will build later this year and start to roll it in mass production of the licensee we work with now for a couple of years.
And these are just some of the examples of the possibilities that you can do with our technology and to use cases that were, quite frankly, not even possible a few years back.
In the next slide , please. Oops. And with that, I'm just gonna kind of bring it to a close and maybe talk a little bit about the company, right? In the end, we are a leader in edge AI technology, again, that provides only the IP, the silicon, the software and solutions. And I'll come back to that.
The market trends are undeniable. The pace of Edge AI, and I think I heard some of the comments from the previous speakers. And if anybody wants to talk about that, we can.
And we're getting strategic traction with companies in these verticals and others right now. Now, before I stop talking about this, and maybe take a few questions and try to keep us back on our time allocation. is that I want to talk a little bit about the left because it's easy to say provide a chip or even provide IP. But one of the things that's really important when you go through a company like us and get the breakthrough performance, you want to make it very easy for people to adopt your technology. So we spend a fair amount of capital working on our software stack in our model zoo to allow people to either bring their models on easily onto our platform, or adopt our models into their own. You can develop your models using industry standard frameworks like tensor flow or pie torch or onyx for those who are AI savvy and put it on. You don't have to learn anything at new to do that because if you've got a bunch of engineers, it's very easy to adopt our technology.
So in the end, we're looking for that breakthrough performance and ease adoption. And with that, I'd like to close my prepared remarks and take some questions, please. Um, yeah, I'll kick it off, Sean, with one question that was emailed in yesterday.
Question - There was some news out of IBM where they paired Akida chips with some IBM products and also some products from Palantir to solve a real world problem for the U.S. Army. So those are really exciting sort of applications. But what does it tell you about the maturity for, for, you know, Akida 1st form, obviously, but also neuromorphic processing, and and about the, um, the verticals, the, the best verticals for BrainChip to generate revenues from.
Sean Answer - Yeah, listen, it's a wonderful use case. It tells me a lot to answer your question about the maturity of it, which is it's very mature and ready for prime time. That's what we're seeing right now. And I could literally use my whole 20 minute allocation mark to talk about that because one thing I would lean into with that, which is so interesting. Most people tend to think of edge AI, even my comments earlier was more discrete, if you will, from the data centre. But with that example, it's a wonderful example where the right tool for the right job. If you read things that were publicly announced, the raw data was kind of overloading and squashing the Palantir system. By putting a Akida to next to those sensors and really turning all this raw data into useful information that was then fed into, it not only allowed it to do it, it allowed it to do it more efficiently and quicker.
And so it's that kind of working relationship instead of just discreetly in some of the examples I had. But this example of working, um, collaboratively with a data centre or centralised services is really where we think this thing is going to have a 2nd set of capabilities in there. And that is a wonderful example.
Now, also specifically about, about, you know, the vertical comment that you made Mark in there too. I'm going to call it a couple more. I mentioned earlier today, just a few moments ago, that we had an engagement with Raytheon that's going incredibly well. Again, a very demanding environment and customer. Um, you know, so there's two.
We also, when we announced the 1500 ship last fall, we quickly announced with a company called Parsons, which is a tier one defence contractor in the in the US that placed an order for several 1000 chips in a multi-year relationship that's going very well, and that will grow in more, many more 1000 chips will call. Another tier one that's putting them into real products and pushing them out to field.
And as recent as last week, there's a company called ForwardEdge ASIC that announced that they have chosen us, not for a particular chip at this point, but for future chips and ForwardEdge ASIC, if you pay much attention to them, is a wholly owned security of Lockheed, a very large tier one defence contractor.
To answer your question about the maturity and the readiness is for real applications, all of those are very demanding companies and we're tier one, so the technology is ready because they're putting it in it.
Thank you, Sean. Some questions online. Uh, yeah, correct. 2 questions online.
Question 1st question for you, Sean, is how would you see his commercial timeline for significant revenue increase?
Sean Answer - Yeah, we believe this year will be significantly more than anything you've seen in the past. And, uh, and we're going to see 26, 27, 28 is the years that we really take off. Why do I believe that?
One is obviously some of the edge market stuff that I had to go very quickly, but I'll take just a moment here, again, cognisant of the time. The use cases, if anybody went to CES, I use that as an example is the best. The number of edge use cases at CES, I would say, were 10 times a year before, and the year before was 10 times.
Not 10%, 10 times. So real products, real things are coming out, real demands.
Um, I've been with this company about 4 years. When I 1st joined most of the conversations were, hey, this is interesting, help me understand it. Maybe I can figure out what to do. Now the conversations I have with customers or like, here's our product plans, help us understand if you can help us do it. So everybody's got the mindset of implementation. So I believe those are the things.
Secondly, what I believe 26, 27 or 28 are our years is the broadening strength of our portfolio. When I said I came here, we had one generation of IP. We have two. We'll exit this year at three. We have chips. Now we have modules now. All the things are some of the previous speakers talked about that people go out and try things and deploy before they make IP decisions. as well as reference designs that we're putting out. You put those 2 factors together, a much stronger product set, a much stronger company, and stronger market forces, we believe this is the year 26, 27, 28 to watch us.
One more question - What part of friendship roadmap are customers responding to most strongly right now? Is it the current capability offerings or what's coming next from BrainChip?
Sean Answer - It's both, and I don't want to kind of give a political answer, but here's here's what happens in technology. I've been in technology for many decades. And this is where you, some of the failures that happen in this industry for, you know, companies or competitors of us or similar spaces. This is the ultimate business to business sale. It is not business to consumers. You're selling to people that you're enabling The customers of BrainChip were enabling their products, which is they're going to sell for revenue, right? So it's a very strategic sale. And they don't buy it on a single point. Yes, it has to solve their problem today with that single point. But they want to see a vision and a strategy 5 and 7 years out. And we provide that. So, yes, their technology today that will solve their problems and that's why we're seeing attraction, but they want to see if we have the vision 5 or 7 years that aligns with them. And so that's the answer to it. We solve it today, and we're also have a roadmap that is compelling for them to say, we can go with them and grow with BrainChip for the subsequent years.
All right, thank you very much, Sean. Do we have any questions in the room here?
If not, then, well, wrap it up, Sean. Thank you very much for your time. Thank you, Mark. and I look forward to seeing you there soon. Thank you very much.
Last edited:
Taproot
Regular
And
this little nugget ?
"BrainChip's Akida has been adopted in prototype airborne sensor systems and the company secured a $12 million defense R&D contract from a Five Eyes intelligence partner in late 2024, validating its technology for classified applications. "
marketintelo.com
Anyone have any more info on what this is about ?
this little nugget ?
"BrainChip's Akida has been adopted in prototype airborne sensor systems and the company secured a $12 million defense R&D contract from a Five Eyes intelligence partner in late 2024, validating its technology for classified applications. "
Neuromorphic Computing for Defense Market Research Report 2034
The neuromorphic computing for defense market was valued at $1.8 billion in 2025 and is projected to reach $9.7 billion by 2034, growing at 20.5% CAGR.
Anyone have any more info on what this is about ?
Sorry my fault! No Brainchip here
#armeverywhere | Arm
More than 50 leading companies are supporting Arm’s evolution into silicon. Across hyperscaler, cloud, silicon, memory, networking, software, system design and manufacturing, this is what it looks like when an ecosystem moves together to accelerate innovation. https://lnkd.in/ehD3sShw...www.linkedin.com
View attachment 96530
Or do you guys think we are involved via partners like intel foundry etc?
Think about a fridge who can cool your stuff during a blackout or something for days without having any other power source… after a massive earthquake or a storm or flood or something.
Or for any device who can run long enough without needing plugging it in etc
HopalongPetrovski
I'm Spartacus!
Thanks jrp173 & Jacob.
Yes. They said at the end they were going to cut the presentation into company sized videos and intended to release them next week.
Unfortunately the guy presenting before Sean went overtime by 10-15 minutes so I think Sean was a little flustered and felt under time pressure.
Still, he did a pretty good job.
Nothing particularly new to my ears but the first question he answered was mine or an amalgam of mine and similar questions.
I asked ...."When do you consider as likely a commercial timeline horizon for significant revenue?"
And.....
As per Jacob's recording........
Question 1st question for you, Sean, is how would you see his commercial timeline for significant revenue increase?
Sean Answer - Yeah, we believe this year will be significantly more than anything you've seen in the past. And, uh, and we're going to see 26, 27, 28 is the years that we really take off. Why do I believe that?
One is obviously some of the edge market stuff that I had to go very quickly, but I'll take just a moment here, again, cognisant of the time. The use cases, if anybody went to CES, I use that as an example is the best. The number of edge use cases at CES, I would say, were 10 times a year before, and the year before was 10 times.
Not 10%, 10 times. So real products, real things are coming out, real demands.
Um, I've been with this company about 4 years. When I 1st joined most of the conversations were, hey, this is interesting, help me understand it. Maybe I can figure out what to do. Now the conversations I have with customers or like, here's our product plans, help us understand if you can help us do it. So everybody's got the mindset of implementation. So I believe those are the things.
Secondly, what I believe 26, 27 or 28 are our years is the broadening strength of our portfolio. When I said I came here, we had one generation of IP. We have two. We'll exit this year at three. We have chips. Now we have modules now. All the things are some of the previous speakers talked about that people go out and try things and deploy before they make IP decisions. as well as reference designs that we're putting out. You put those 2 factors together, a much stronger product set, a much stronger company, and stronger market forces, we believe this is the year 26, 27, 28 to watch us.
Hi Hop,
I was also on the live feed, enjoyed listening to Coby then switched off until Sean appeared, I felt he was a little rushed, though nobody
was insisting he speed through his presentation, he came across solid and assertive, has reconfirmed the company's view of increased
revenue slowly increasing 2026 through 2028, which really ties in with how things are currently tracking...AKD 3.0 by years end, further
extending our runway into the 2030's in my opinion.
I remain positive, pity my holding has been reduced through financial pressures, but intend hanging on as long as possible.
Kind regards......Tech (Chris)
100 staff was a cashflow break even metric that Sean presented to the board !
If still relevant , we could be in for share price re-rate during 2026 !
HopalongPetrovski
I'm Spartacus!
Hi Tech.Hi Hop,
I was also on the live feed, enjoyed listening to Coby then switched off until Sean appeared, I felt he was a little rushed, though nobody
was insisting he speed through his presentation, he came across solid and assertive, has reconfirmed the company's view of increased
revenue slowly increasing 2026 through 2028, which really ties in with how things are currently tracking...AKD 3.0 by years end, further
extending our runway into the 2030's in my opinion.
I remain positive, pity my holding has been reduced through financial pressures, but intend hanging on as long as possible.
Kind regards......Tech (Chris)
I am in a similar boat, having had to sell some down to stay afloat which really sucks, as we were clever enough to recognise the potential early, (when it was more expensively valued)
, but, because its taking so much longer than anticipated, forced to sell down at these miserable prices. "Oi vie"....whatchyagonnado? Live, laugh, learn......I guess that's what we're here for.
Sean (imo) isn't a natural "presenter" but has gotten a little better over the years.
Unfortunately being put under stress doesn't help anyone and at the beginning there was a moment or two when he looked a little bit like a deer in the headlights.
But though he recovered, you could see he was aware of the clock and I think he felt compelled to "be finished" by a certain time, which put him under undue extra pressure and I think it showed.
But, like I said, I think he is looking more and more confident and satisfied with progress behind the scenes which of course none of us are properly aware of.
Apart from your's, think I also saw Ken Scarince's name watching on along with Steve Liebskind who had a crack at a non authorised directorship last AGM.
Just mention out of interest.
Our company certainly has developed a somewhat cult like following.
Shirley we can't all be wrong???
HopalongPetrovski
I'm Spartacus!
HopalongPetrovski
I'm Spartacus!
"They looked Iranian": Security footage shoes Trump negotiating........
Similar threads
