Hi, everyone. Thanks for joining the session. I'm Kingsley Crane, a technology analyst here at Canaccord. Really pleased to have the GSI technology team with us here today. GDA, thanks for joining.

Thanks for having me.

Let's kick it off. You just reported your June quarter. You've described this as a pivotal point. What does GSI look like today versus 18 months ago and just at a high level?

Yeah, so we certainly progressed the company quite a long ways. 18 months ago we were just coming out with Gemini 2 which is really our first commercialized part for the AI market fast forward to now since then the Gemini 2 is in production now as far as the hardware is concerned we've gotten some third-party validation from folks like Cornell Cornell University actually had a board of ours, and did a RAG comparison with a NVIDIA GPU, and it found at comparable performance, we were 98% less power, so it was certainly a huge movement in the market space. Since then, we've also engaged in a few POCs. We have one drone surveillance POC that is being funded by the DoD. We also recently won a phase one POC out of a municipality in Taiwan for smart city which is nice because it leveraged a lot of what we're doing for the POC for the drone so it's not a complete new lift. Since then we've also gone quite a ways with developing a AI SDK or AI assisted SDK I should say and this is important because it's going to help us enable the ecosystem. If you look at our model now, we are writing our own applications for all these POCs, which is fine to really showcase the technology, but to really scale, we're going to have to have tools that the customers can use, and so we're quite a ways with that SDK. We'll have our alpha version out in this fall. We see what else have we done since then. We've won a couple SBIRs, one with the US Army for a recognized edge node. This could be a nice product for us and for the Army. Essentially, it's gonna be a server that can do object detection or it can do SAR imagery, and it can be done, again, at the edge. So it's a recognized server they can put at the back of a Humvee or something. Let me see, 18 months. Oh, we've also started our next generation device, the Play-Doh. Plato is going to be, you know, it leverages some of the technology, obviously, from Gemini 2, but it's going to adjust a different market. So the way that Gemini 2 is, is that it's, you know, we have a small bandwidth coming from memory into our chip, because really the intent of the chip is to download a model or database one time, and then once it's in our chip, we run it really, really fast. So our internal bandwidth is extreme. So it's perfect for search applications or HPC kind of applications. Plato on the other hand is gonna be developed more for LLMs at the edge. So we're gonna open up the pipe so that we can get data into the part faster. And then we'll scale down the internal bandwidth to kind of match that. And so that'll actually be used for vision language models, large language models at the edge. And it's gonna be extreme because it's gonna have almost data center performance at a two to 10 watt power budget. So it's gonna be a powerhouse for edge applications.

Really helpful overview. So investors are trying to target, are trying to understand the AI supply chain and target the bottleneck. Can you just help us understand why SRAM is so important for AI, how durable that business is and what is a secure AI super cycle at this time?

Sure. I'm going to ask that in a couple of different ways. So first of all, it's important to our company because right now it's the cash cow, right? And so we're just starting the AI story. And so we need something to offset the bills. And we've been doing SRAMs now for 30 years, shipped over 140 million devices. So we're certainly a leader in that market. And so it's really helped offset the bills or the cost there. As far as the market itself, SRAM isn't directly in the AI. It's not in a data center, but it really helps the infrastructure. And what I mean by that is one of our largest customers we've talked about is Cadence. They make emulation systems. These are systems that if you're an IC manufacturer or designer, I should say, you do a design. And in the old days, you just do a simulation. And is the part working? Do I have any major bugs? And it wasn't until you went to FirstSilicon, got the actual chip, you looked at it and said, wow, it's dead on arrival, part doesn't work. That's a real problem nowadays because you have to create a math set in order to get that chip. And the math sets cost upwards of $30 million a math set. And so you don't want to be wasting a lot of $30 million math sets. And so now guys like Kanan's make these emulation systems where they actually emulate the design in software. And these are large, expensive systems. and our highest end parts go into those systems. So we're really helping the front end design. Another one of our larger customers is KYEC. They actually are part of the manufacturing process. In this particular case, they do burn-in. Burn-in is basically a way to electrically really challenge a part to make sure there's what's called no infant mortality rates. In other words, you don't want a part to be put in a system, go out to the field, and then fill because it's a weak chip. And so they do burn it to try and get rid of those. And KYC is doing the back end for all the latest GPUs that are out there. And they use one of our high end parts. So we're kind of indirectly supporting the AI with our SRAM division.

Can you help us understand why compute in memory is so critical from a performance and cost perspective?

So there's this thing called von Neumann model. I don't want to get too much into detail on it. But essentially, if you look at a way a GPU and a CPU work, they have their processing elements. And so when they're being asked to do something, do a calculation or some kind of process, they have to go outside the chip to fetch data from memory, bring it back, use it. And once they've used it, they have to write it back to memory. So there's this constant data transfer, data flow back and forth. It takes a tremendous amount of power, which I'm sure you've all heard. You know, data centers, what's their biggest problem? So with the APU, we've done something much different, and this is all patent-protected because we know we have something unique, is with the Gemini 2 family, as I mentioned, we have a small pipeline going into the chip. We bring in data one time, and once it's there, we run it very fast. And the way we do that is we have, first of all, a large memory inside, but the process or the calculation is actually done in the memory bit line, in the memory itself. So our bit processors are actually coupled with our memory. So we're not going out fetching data, we're not bringing it back, we don't have this constant transfer. And so the performance is high, but the power is extremely low. As I mentioned, this Cornell paper, I mean, 98% less power, and that's because of that CIM architecture.

So for Jim and I, too, can you help us get a sense of that path towards commercialization, just moving from proof of concept into product? And then what do you think success looks like in a year from now?

So as I mentioned, Gemini 2, the hardware is production ready. The software is catching up right now. I mentioned we're going to have that AI-assisted SDK in the fall. In the meantime, we have two engagements on POC, as I mentioned. One is the drone surveillance. The other one is a smart city out of Taiwan. we're writing all of the applications for that ourselves right now. And that's kind of our model. I mean, we're a smaller company and so we have limited resources. So we're gonna write an application for a certain market, show that we have a good solution and advantageous solution. And then at that point, that coupled with the AI assisted SDK will allow us to then use system integrators for each of these markets to address the broader market. So POC for the initial and then the system integrators with our tools for the launch. And so we're looking at the POCs, the two we have identified now will last through this year. We have a couple more that we haven't announced yet but we're still working on that'll take us into next year. And so what we're looking at is completing the POCs between the end of this year and next year and having early production by the end of 2027.

To what extent do you feel like that those engagements that you're having right now are reusable?

Great question. They are very reusable. So it took us a while to do the first POC with the surveillance drone. They had a requirement that was called time to first token, which is essentially you have a video feed coming in and, you know, there's events that happen and you need to identify, is this an issue? Is this a bad event? And you need to do that very quickly. In this case, for the drone, it had to be less than three seconds. And then at that point, once it identifies, you know, okay, I have a problem. What is the problem? You have to identify. It's, you know, a truck that went through a gate, you know, unannounced. And then at that point, you need a response. You know, what's the recommendation? What do you do? You know, most of this particular sentinel program it's usually military bases or government buildings the response might be take out the truck we don't know what the response is but we have to at least do a recommendation now the work we did there a lot of it's reusable for the smart city smart city now is um you know somebody jogging in a park and clearly grabs this chest and goes down the system has to identify this is not somebody who's just tired and wants to take a nap this is somebody who's having a a medical emergency, pass the identification. The response is, alert EMT, car has an accident. Well, that happens all the time, but car starts to smoke, catches on fire. It gives you an alert to the fire department, you have a car on fire and here's where it's located. And so that going also to other POCs will have similar, other POCs we're looking at are like industrial inspection, tying the first token, very important. So a lot of what we use from the last POC we can use to forward future ones.

So when you think about a software developer kit or SDK, building mindshare is really important. And then if you move towards an AI-assisted SDK, it could further increase or improve time to value for customers. Can you just double-click on why that's so important for GSI?

Yeah, it builds the ecosystem, right? As I mentioned, we're a small resource company. we don't have thousands of software engineers that are writing applications for us and that's just not gonna happen for us. And so we need to enable the market for them to write their own, whether it's the customer directly or whether it's some of these system enablers I mentioned. And what this SDK does is it allows them to write at a higher level language and have it be translated into a language that our part understands. To give you an idea, the drone POC we did, The application we wrote there, it was before the AI-assisted SDK. It took us about a man year to write that. With this AI SDK, which my internal team is starting to use already now, it's going to take that down to weeks. So it's going to really accelerate the time to be able to get out these applications and also allow customers and integrators to use them without understanding really the machine code of our part.

But maybe it would be helpful, we've talked about Gemini 2, but just a reminder of why Gemini 2 is so important, why it's winning, and then the next step that Plato takes on top of that, more importantly.

So let me give you a real-life example of why it's important. I keep talking about this drone sentinel POC. So originally, the drone manufacturer was going to use NVIDIA. I mean, everybody who knows NVIDIA use NVIDIA, right? And there was two critical components to this POC. Number one is time to first token, I mentioned, of three seconds. The other one was they needed a power budget of less than 50 watts. I mean, the drone needs to go up, and it needs to stay in flight for some amount of time. And so they looked at NVIDIA. NVIDIA gave them the three seconds time to first token, but their power was 160 watts. So it was over three times beyond their budget. So they looked at Qualcomm Snapdragon. Snapdragon. Snapdragon gave them the sub-50 watt power, but their time to first token was 12 seconds. I mean, it's four times the allowable amount. I mean, assuming that the reaction time is three and it takes you four times longer, you know, a lot of times the danger is escaped or what have you. So they had to get the critical. And that's when they looked at GSI and we gave them sub-50. In fact, we're at 30 watts of power. And right off the bat, we were three seconds time to first token. Just past June, we actually did a lab demo, and I say we, us and G2 Tech, who's our drone partner, did a lab demo for the DoD, and the time to first token actually came in at 2.5 seconds. So we've exceeded the 50 and the three seconds.

So maybe talk to us more about the balance sheet, over 70 million in cash, no debt.

How much do you view that as an asset for you and then just thinking about that enabling the roadmap sure yeah it's critical to have that so as you mentioned 77 million no debt uh you know we have the sram division that's offsetting a lot of the bills obviously but we're still burning some cash right now until the until the apu takes off and so we're burning roughly four million a quarter so we'll say we'll say 16 million a year um and and that's and that's flat except for we will have a slightly higher expense in the the spring quarter when we have the tape out for the Play-Doh. And that'll be an extra couple million dollars. I mean, we're not using the latest advanced technology there, so it's a little cheaper mass set. But besides that, the 77 million clearly is enough to satisfy all of our near-term goals that we need and milestones that we need to cover. So yeah, we feel that we're in great shape there.

Just to complete that funding picture, how about SBIRs or other sources of non-dilutive funding from the government?

So SBIRs, they help on the funding, absolutely. So we use SBIRs. If you're not familiar with them, it stands for Small Business Innovation Research, I think, is what it stands for. Anyways, it's basically a way for the U.S. government, through one of the DoD elements, to help fund some technology for smaller companies. And so we use that money as an offset to R&D costs. But for us, it's also a way to engage with these entities. We've won two SBIRs with the Air Force Research Lab. We've won two with the Space Development Agency, and we won one with the U.S. Army. So we have two that are still active. One is with the Space Development Agency, and that was to fund radiation testing on our Gemini 2 part. And so what the SBIR is doing is taking a commercial off-the-shelf Gemini 2 part, and we're doing radiation testing, looking for any kind of single-event latch-ups, which is very bad. And we're also doing a total ionization dose testing, which is basically how many ions can the part absorb before it stops working. So we actually did the radiation testing in June. we're still waiting for the report but and because it's the the the testing is done by a third party but we did have a gsi employee present at the at the um testing and there was zero single event lash-ups which is critical which was fantastic and again this is on a non-optimized commercial part and then at the end of this month or beginning of september we'll do the the the tid um testing and we are confident we'll do well there because our srams have always done well and we have similar technology. And so once we have those two elements in place, it's going to help enable the DoD entities to be able to have a part that they can use in space. Any of you familiar with some of the other technologies like GPUs, they inherently are not good for space applications. And so right now, if you look at some of the applications, there might be a LEO satellite right now that has sensors that are basically taking the data and sending it to Earth for image creation or what have you. They want to be able to do this in space. In other words, the satellites take the sensors, and then we can create the images on the satellite. And then now we can send that to Earth saying, hey, these guys are deploying equipment or trucks or something or deforestation or what have you, whatever they're looking for. And so, you know, those SBIRs are critical not only for offsetting R&D costs, but also for, you know, creating, you know, new markets. One last thing I want to touch upon was on the US Army. I kind of mentioned it earlier, it's for a ruggedized edge server, but this is something that could be productized by us for the DoD, and so this is not only a way to offset R&D, but also a potential revenue stream for us in the future.

Right, I think Elon would love to put a bunch of GPUs in space, but we'll see how quickly he can do that. In terms of needing power efficient AI, of course we need in the data centers, but we need it at the edge. It's so critical there. And so thinking about use cases in industrial automation or smart city is really compelling. You talked a lot about defense. Is defense the beachhead market where you break in and then maybe just talk more about opportunities outside of that?

Yeah, I think it is only because those guys have shown their early interests and they've made commitments. These SBIR dollars, I mean, they're millions of dollars. And so they're actually making a financial commitment that they're saying our technology is something that is going to be useful to them. And so I do see that the early wins and interests we have come from that market. So, yeah, I do see that that would be where some of our early revenue would come from.

What KPIs should investors be watching in the business over the next couple of quarters to really judge this inflection?

Sure. Let's close out the two POCs we have now. as I mentioned the drone one we've actually candidly on our part have have finished all of our deliverables the the drone manufacturer now has to finish theirs for the final field demo but then we have to see how that you know the final demo goes we have to close out the phase one POC for the smart city at which point it's going to go to phase two so phase one is basically so let me take a step back this municipality already has cameras installed all over the place and what they do now is they just record that's all that's all they do and so if there's an event that somebody is interested in they have to go back to the footage and go back through old footage to see what happened and what this municipality wants is they want to have you know intelligence up front and so they want to understand is there an event happening right now what's the event and how do we respond and so phase one is to take video from this system for 20 cameras and go through, in fact, we have a small demo on our website you can look at. It's one of the demos that we use to win the POC, but it kind of gives you a feel for what it does. Phase two, we'll bring it up to 80 cameras, and we'll also now include audio. So the initial phase one is only video, now it's audio, because some of the issues, they're going to be having some of these at schools and they need to understand if there's any abuse or anything happening. And abuse isn't always physical. Sometimes it's verbal. And so they need to be able to see that. And then phase three, by the way, phase one, we'll have our deliverables in November. Phase two, we'll start about then. And then assuming we win phase two and we win phase three, that is the production deployment that would happen sometime at the end at 2027, and that could be anywhere between 2,000 cameras and 6,000 cameras. And to give you a feel for what that means to us, there's one Gemini 2 chip for every four cameras. So it'll be a significant hardware sell for us. And then on top of that, we have an annual recurring license to the application itself to keep it running. So it'll be recurring revenue there. So we have the POCs we need to close out. So another KPI is obviously the AI-assisted SDK will have the alpha release in fall for certain customers. And they're going to try and break it and get any bugs out of it. And so our intent is to release it sometime next year to the mass public. Other KPI, obviously the tape out or the end of the design of Play-Doh. You know, Play-Doh is going to be real important, as I mentioned, because it's going to be an LLM powerhouse at a very, very low wattage. And so certainly in the springtime, you know, we need to get that out.

So just even thinking about centralized AI, we have so much more room to use more of that. Decentralized, we have even probably a longer runway for that to permeate through all these different form factors. How do you, as a company of your size, compete against some of the larger players? How do we make sure that the best technology does win?

Sure, yeah, so let me take, I'm gonna take one step back before I answer that question. You have the Qualcomm's, the AMD's, and the NVIDIA's are huge companies, and then most of the other AI companies are very small, and they have an idea. So we're more of an AI startup as far as the technology goes, but we're not a semiconductor startup. And so that's one thing that people need to understand. And as I mentioned earlier, we sold 140 million SRAMs over our 30-year career. And we did it using TSMC as our fab, using ASCK as our assembly house. Fast forward, our APU, our Gemini, and our Play-Doh will use TSMC and ASCK as our two assembly houses. So we have a long 30-year relationship with them. We already have a full operational team. So as far as ramping, that's not going to be a problem. As far as how do we win, we bring both performance and low power, and that's critical. As I mentioned, on a compute basis, NVIDIA is strong. They have good performance. Their power is high. Guys like Qualcomm, their power is good. Their performance is low. And so there's going to be a lot of applications where you're going to need the performance and the low power. And so on a performance per watt basis, that's how we beat those guys.

So we're close on time, just want to make sure the audience has a chance to ask if they'd like. So you've talked about the technology, and it's been validated, and we're starting to see an inflection there. But what else do you need to prove either to customers or to the market over the next 24 months?

I'm not sure proof is the right word as much as execution. We have to get the software tools out. The hardware, as I mentioned, Gemini 2 production ready right now, and Play-Doh will be out next year. So it's really getting the software tools in the hands of the customers and the market so that we can really create that ecosystem. So it's really more of an execution than it is. I think we've proven that the technology is real. The Cornell paper helps and other benchmarking we've done, winning the bake-off for the Sentinel program. I think we've done the proof part. Now, we just have to do the execution on the software side.

E.D., thanks so much for joining us. Looking forward to seeing that execution.