Weebit Nano Limited (WBT) Earnings Call Transcript & Summary

Good afternoon, and welcome to the Weebit First Quarter 2026 Investor Update. My name is Danny Younis, and I handle Investor Relations for Weebit. With me today, we have the CEO of Weebit, Coby Hanoch; and for the first time, the VP of Marketing and Business Development, Eran Briman. Before I hand over to Coby, just a reminder, we will be having a Q&A session today. If you have any questions, if you're in the room, just put your hand up or if you're online, just type them into the Q&A box that you see at the bottom of your screen. We're expecting a lot of questions, but we've provided enough time, I think, to go through most of those questions. We've got roughly an hour, but we can run a little bit over if we need to. I would now like to hand the webinar over to Coby.

Hi, everyone. To me, so good to be here, as always. I guess we came here because of the Semiconductor Australia conference that was yesterday. I think it was a very good conference. It was well attended. I don't know if anyone here -- was anyone at the conference? It was really -- I think there were like roughly 300 people or so. This time, we had a booth there. So you see the sign from Semiconductor Australia up here where we're using it kind of and show the demo. I guess we'll be bringing the demo to the AGM, and we'll be showing it there as well. So we'll be back -- I'll be back in a month. And yes, I guess, let me share the screen again with this. So just a few slides, not really much. I think everyone knows, knows Weebit and knows what's going on. So, okay. So we'll just go through the slides like this. So I just thought I'd show you an update. I found this very interesting how just a year ago, the prediction of the size of the semiconductor market was the blue line here and people were saying, yes, it's going to hit USD 1 trillion by 2030 or a little bit later and stuff. And now with AI just going wild in just one year, we see analysts already updating their predictions and the size of semiconductor crossing the $1 trillion mark already in 2028 and so on. So it's -- again, every time I come to Australia, I have this challenge of trying to explain to people how big this market is, how important it is. You guys know, but like yesterday at the conference, even though a lot of people knew what semiconductors are, we still had to emphasize it and of course, with other institutions and other people that we meet here. So I think this really shows you what's going on. It's crazy, the level of investments in this market, I always repeat it because the numbers are just mind-boggling. Every time you see it more investments and more strategic deals, Intel now -- on the one hand, Intel is struggling. On the other hand, they have a new CEO who's really a magician. He's amazing. I know the guy, and he's just capable of doing amazing things. And now he's managing to get investments from the U.S. government, from NVIDIA stuff. India now is starting to invest a lot. They're setting up foundries in India. In terms -- I'm sorry. I need to do this. In terms of the ReRAM market itself, that every once in a while, I refer to the old slides, and it's interesting to see how year after year, the numbers evolve. I think the interesting thing here is to see they are talking about ReRAM being today about 27% of the market. And I think this is already what we've been talking about for a long time. ReRAM is being accepted now as the new standard. And you can see how the ReRAM market share is growing quite a bit, actually more than doubling or just about doubling in just a few years. But also look at the size of the market, it's like 20x. So not only is the market share growing double, but it's 20x the market. So it's actually ReRAM section is 40x in just four years. And that's a huge growth. That's a huge growth. Right now, the companies who are addressing this are basically TSMC and Weebit. The companies that are commercializing the ReRAM that actually have customers that people are buying it, it's TSMC and Weebit. UMC has a ReRAM that's been qualified, but it's been qualified for years, and it's only at 105 degrees, which is really not accepted in the industry as high enough. And to the best of our knowledge, they don't have any customers. Even the owner of the ReRAM actually, Nuvoton is using TSMC. And there's a new entrant. We all knew that GF is looking for ReRAM. We were trying to be the ones who get in there. Their internal R&D has been pushing strongly that they make it happen and the management gave them a chance. So they announced that they have a ReRAM. I think the good news is you can see that people see ReRAM is really needed. They haven't qualified it yet in their announcement, mentioned automotive. I think they still have some way to go. I personally think they made a big mistake because they just don't have the resources to maintain this. At a certain point, it will have to collapse. That's my personal belief. And so I think both UMC and GlobalFoundries are definitely high on my potential customer list still. I still believe TSMC is going to be a challenge even though they are on my customer list, and I believe eventually we can get there. But this is kind of showing you the market. Notice. You can see, by the way, the split between different markets, analog, MCU and ASIC. And you can see that the initial users are the analog guys. I think you guys already know, we started talking more and more about analog. onsemi is basically an analog company. Even with DB HiTek, it's a BCD process, which is mostly for analog companies. And you can see the adoption. It's no surprise that Weebit's first fab customers are analog. But then after that, the MCU market starts going up. And then after that, you have the ASIC and SoC margins going up. So this really shows you, by the way, that it's reflecting also what Weebit is doing. The fact that we signed up with analog companies. I think that's part of the thing that Yole has been looking at and saying, okay, we can see the money coming in. I mean, you already saw the revenue number for the fiscal year, the $4.4 million. And I guess now we're very happy with the recent announcement of the quarterly that -- this quarter, we got $7.3 million from customers. Now I think it's important for me because I know some people were really getting excited and oh my God, it's already cash positive and look at these numbers and everything. It's important for me, and I like to be honest with things. We're getting paid based on achieving milestones. And sometimes some quarters have more milestones, some quarters have less or whatever. And some of the milestones are bigger and bigger payments and some milestones are smaller. So I want people to understand it's great, and we're obviously very proud and very happy of the money that came in. I just don't want people to think, okay, now that's how it's going to keep growing, and we're going to see record quarters every quarter and whatever and extrapolate stuff. So we need to keep it in proportion. It's good. I mean you can obviously see from these numbers that the revenue numbers, I guess, for this year are going to also grow somewhat. And we're not giving any guidance, of course. But that's -- I think this is a good map of what is happening, and you can see how the Yole numbers are changing and growing. And Yole is finally accepting. Yole is close to STMicro, which is which is the only PCM provider actually. And they always -- at least in our mind, they kind of tend to grow the PCM market just because of that influence from STMicro. It's the only supplier of PCM. But anyway, that's kind of -- again, I need to click on this thing. So, recent highlights. You guys know everything here, the AEC-Q100, that's already old news for you guys. Record revenue for the fiscal year, then taped out at onsemi. That was a -- you can see onsemi has a team of I estimate at least 30 people, almost full time working with us. It's a big team. It's a big effort. They are really focused on pushing this forward as fast as possible. It is strategic for them, and it's great -- it's just a great cooperation. It's -- you can see with IDMs, it's really great because they really want to get this out and they want to get their products already moving and they want to. So it's really been a great partnership and this tape-out, and we now have to sit and wait for the wafers, and we'll see how long that goes. We, obviously, we just talked about the record quarterly customer payments. And obviously, that's nice to see there. And by the way, I want to point out, I mean, people are looking also at, oh, this was a positive quarter and stuff. You need to note that there was also the yearly refund payment from the French government that they gave us. So I'd just like to be as transparent and as clear as possible on things. And then you know that we have a target of three fabs and three product companies to sign up this year. I'm very glad that we actually achieved the three product companies already, and that was really nice. We started announcing the first one, and then we have two more now that are -- so there are three products that are already being designed with our ReRAM in them, embedded in them. Obviously, again, the design takes time and they need to manufacture the prototype wafers and test and see and qualify their products. But the clock starts ticking towards revenues eventually. So that's kind of the update. I always like to point to why Weebit is actually managing to make this progress. What is the difference between Weebit and all these other companies that are trying to build ReRAM or work in this nonvolatile domain. And I really think it's this combination of we have the company that is -- I think we're strong. By the way, the cash position is obviously something that is very important. We have $91 million in the bank, and that really enables totally focused on let's improve the technology, let's get the business, let's close these deals. I'm not worrying or spending time on how am I going to raise money now, and that's very important for us. We have money. Again, you start doing extrapolation in fiscal year '25, the spend was about $25 million roughly. So you can see we're feeling good about it for several years now and that we can totally be focused on getting the business, closing deals and in parallel, growing -- improving the technology. The people, you guys know our Board, which I'm extremely proud of each and every one of them. But beyond that, the company now is about 50 people already. We've grown. Obviously, we need to start supporting more projects and these things. We do have roughly 1/3 of the R&D is PhDs in physics and chemistry. I mean we really have a very strong team. You look at the VPs, I think the average experience is over 30 years in this industry, very experienced, very well-known people in the industry worldwide. I'm not talking about just in Israel and even the levels below, I mean, the average age in Weebit is much higher than you see in normal high-tech companies. And it reflects that you need that level of experience. You need that level of know-how of these things. So, really, that's very important. And Leti, of course, is supporting us. So I think it's great to have that. Obviously, we have a very good technology. It's part of it is all of this R&D that is going in with very smart people. We managed to make the technology work and I can tell you, we've hit endless obstacles along the way, but what's important is that you have the team that knows how to overcome them and move forward and fix things. And so really amazing technology, AEC-Q100 and all and so on. And we're providing the solutions now for the customers. And there's actually -- you have onsemi and the analog guys, and we're working a lot with analog guys. There's quite a bit of activity going on now with companies that want edge AI and want to see how we bring in ReRAM there, and there's activity around there and all these other things. And then it's really the customers. We have a very good relationship with the customers. Yes, some of them are just still hesitant. There still is that perceived risk. They still haven't seen this in mass production and so on. So each one at their own pace, but we are pushing forward. We are -- we still have to sign two more agreements. We're working hard on this, trying to get them in this year, and we don't have a lot of time. I remember last year, at this time, I was sitting and people were saying, you don't have time left. You said you're going to have a deal, you don't have time left, and we got it done at the last minute. We're pushing as hard as we can to get this done this year. We are engaged with a lot of companies with a lot of fabs, okay? I mean it's just fabs. I think it's definitely more than 10 that we're engaged right now in discussions, in evaluations in all of this stuff and each one of them with their own story, why and whatever. And -- but there is a very good -- we're in a good place there. And onsemi definitely helped having that deal. And I believe that when we sign another one or two, it's really going to start -- I don't want to say, opening the floodgates, but it's going to get us moving forward. So, and the customers, by the way, the product companies, they -- we have a huge advantage. We have a huge advantage in the design side. People don't realize. We talk about the PhDs and physics and chemistry and stuff. But we have an amazing design team, which is recognized in the industry to the point where sometimes people who try to develop ReRAM ask us, can you guys take our ReRAM bit and do the design around it because we don't know how to do that. And so we have an amazing design team. Customers realize and these product companies, obviously, our goal is to eventually have enough standard modules that people will just come and use them, and we won't need to do manual work there. I mean the goal is to have high margins and then not be doing services. But it's -- the customers know that if they need that support, if they need that design service, we can give it to them. And that's a critical thing. And especially the first customers, they always need some more handholding and stuff. So we have that team. I strongly believe that when you build the brand name that with Weebit, you're going to be successful. That's what really gets the market going. If you look -- if anyone checks my history and the different companies that I was in, almost all my life, I've been competing with companies that gave competing technology for free, literally for free. They were big players. The customers would buy these huge packages from them and then they would tag on, on top of that for free, the competing product. And I had to go and compete with that. And the only way I could do it was just give good service. The customers knew. And I would meet the CEO and I would look him in the eye and I would say with the companies that I was at Verisity or whatever, we're going to -- I'm telling you, no matter what happens, you have a problem, you call me up, I'm bringing in the cavalry. We're coming in, you will be successful with our product. And I was -- we were selling Verisity was selling more than $100 million per year when Synopsys gave its competing technology, which was good for free. And we got more than $100 million a year because people value good service. And that's really what we're doing here. I go and I look at the CEOs in the eye and I tell them, with Weebit, our team will make you successful. We are there. We have a very experienced team, and they are there for you, and we will make sure that your product won't fail because of us. It might fail because of you, but it's not going to fail because of us. And that's a key thing. So that's really just to give you guys where we are, what the status is, if I'll stop sharing, and we can probably go to Q&A side.

Yes. Thank you, Coby. So we will now move to the Q&A part of the session. Once again, for all those online, and there's quite a few of you, if you have any questions, please type it into the Q&A box at the bottom of your screen. I can see they're starting to come through. But I think just to start off with, we might start in the room. Are there any questions from any of the attendees in the room.

Question on next year. We obviously -- I mean, you're too early into the commercialization journey to be giving earnings guidance, as you said. This year, I think it was really helpful for the investor community when you issued those targets around the product customers and the fabs. Is it relevant to have a similar target or goal that the investor community can look at and measure against next year, given we don't have that earnings guidance status, if you like, yet? Is that something you've considered?

This guy is on my case saying, what are you going to announce on the AGM? I mean during this trip, how many times have you asked me that question, we're thinking about what we can do. It's very difficult to actually -- I mean, the targets that we set for this year were, a, very aggressive; and b, I mean there's a big risk on achieving these kinds of targets. And it's very hard, especially when you start growing to keep saying how many customers do you expect to have. And I think it's going to be hard for us to talk about number of customers, but we know you guys need some sort of targets or guideline or whatever, and we trying to figure out what would be the most meaningful goal for you and to keep going. So we're discussing this, and I guess we have a month until the AGM, and we'll figure out what we want to present there. By the way, can you mention the names? I'd like to get to -- David, I know.

Just regarding these -- the deals that you've made with the product companies, can you give me an indication of the scope what one may entail? Is it just particularly one particular model of chip that they're looking at? And if it is one, how much extra negotiation or whatever that has to go on if they decide to do another one or expand what's existing?

I think we have you here.

Well, I think in the quarterly activity report, we mentioned also the specific markets that these guys are targeting. Battery management was one of those and then some security-related applications and products. These deals, they differ from one another. It's not one deal that fits all of them. In most cases, they get a license to use the technology in a specific product, while others would like to have this in a complete product line. So, entire products. But I think at this stage, we won't be able to give any specific details about these.

I guess what I can say is right now, they're mostly coming with one product. They want to test the waters, right? I mean nobody jumps head first to the deep end without knowing enough. So the deals that we have now, I guess, basically are, again, I can't talk about all of them. But in general, let me talk generality a bit. Most of the companies that we talked to and we're already engaged, obviously, with more product companies. Most of them are basically thinking of one first product. And we have that negotiation. Some of them want some NRE work and stuff, others are actually willing to use the more standard module. Now obviously, once they do one product and they feel comfortable, they'll want to do more. So it's -- we look at it as we're opening the door to that customer, and now we want to go and expand. So now once you sign the first license agreement, normally, you can just add addendums and it becomes much easier. And that's -- again, that as you grow, the overheads start shrinking because a lot of things you did already and now you can. So the customers you already have agreements with, you normally just add an addendum and you specify the new product. And I mean, if it's a different manufacturing node, there are a lot of parameters, by the way, in these license agreements that you define. And the license fee to give everyone a feel for this thing, the license fee varies by 1 million different parameters, by the size of the memory, by the node that they're using, if it's 130 or 65 or if in the future, it will be smaller nodes, it kind of changes. So there are a lot of parameters. It's hard to give a guidance on, okay, this is what it's going to look like. But the key thing is once the company starts working with normally, they will -- they get to know your technology, their designers get to know it. And then they can just start using it more and more and more products in parallel and you start having a lot of addendums to that license.

In November last year, you indicated that once you felt once you had two or three customers that FOMO would sit in. And you've also talked about in the past, crossing the chaser. So, on those metrics, where do you consider you are now?

So we were just talking about it this morning. We're right now in that crossing the chasm bowling. We're at the point where DB HiTek helped us drop the pin of onsemi, onsemi definitely now got others moving. I think we're getting closer to that tornado. I think that, I don't know, one, two more agreements, we'll see it going. So it's -- we're sensing it. We're sensing the fact that, I mean, we're talking to all of these foundries, IDMs, et cetera, and people know they need ReRAM. I mean, by now, TSMC has it in mass production. They need to compete. Now that GlobalFoundries announced it, it got more attention. The market is moving. It's moving at semiconductor pace, but it's definitely moving.

Yes. Last one for me. I was just interested in understanding the difference between the fab and the IDM. Obviously, with DB HiTek, we know you've got to go through the tape-out, the wafer production, qualification, et cetera, and then start talking maybe some in parallel, but then start talking to customers about their design and then going through that process to get to a customer chip. With the IDM because they are their own customer, is it safe for us to assume that, therefore, the time to a product and market is shorter with the IDM because you don't have that secondary design, test, qualify process?

The answer is yes. The answer is yes because the thing is when you're working with two independent companies, there's that trust level that I mean the product company, the issue is how much they trust the fab or the foundry in this case, that everything will go well and it will be ready and whatever. So they tend to want to wait until qualification. Now sometimes, and you can obviously see companies actually end up signing up before qualification because they can already see the qualification happening and stuff. But it still takes time. I mean there is still -- it's two different companies, and they need that trust level. The thing in an IDM is it's one company. When it decides to sign up to manufacture, it's because it knows that it wants it for its own products, okay? So there's -- you kind of skip that phase of convincing because the only reason why they sign a manufacturing license is because they know they want it for their end products. I mean, otherwise, why would they sign up with you. So, in that sense, that doesn't exist and you can parallelize more. Now where and how and what's happening onsemi has been very strict with us on don't talk about anything. So we can't go into specific what's happening with onsemi. But in general, you're absolutely right.

What's your confidence? I'm sorry, my name is [indiscernible]

Okay. I just -- normally I meet you every time, so I need to know.

What's your confidence level in achieving or in signing the two remaining IDM fabs and the qualification of the South Korean company, DB HiTek?

So, qualification is actually moving. We had some....

Sorry, if I may just add by 31st December.

31st of December, yes, of course. That I understood very well. So, with DB HiTek, we had some hiccups on the way. I mean, always expected that you have some issues, but we are continuing. We believe that the qualification will be done by December 31. Right now, whatever big surprise happens. But in general, we're on that path. With the fabs, we are in advanced stages with one, with others. We're also moving forward. We're really trying to get more done by the end of the year. I guess I can say some might slip into 2026. I can't -- the 31st -- the problem is it takes two to tango, right? And these guys sometimes drag things. So we believe we can definitely close -- one of them we can definitely close. The second one, we are really pushing hard. And actually, there's, again, we're engaged with a lot of them. I don't even know who -- sometimes who it will be because we're talking to several. But the goal is really to push as much as we can to have these things done by the end of the year, and we'll see. It's work in progress.

The equity incentives of the other directors are hang on.

Well, my equity incentive, believe me, I'm thinking about that a lot. Believe me, I'm thinking about that a lot, mine and his. We all -- our equity is tied to this, and we're really pushing as hard as we can to get it done. You can imagine.

Wishing you the best.

Are there any questions still inside the room? We can revert back to you at the end, okay? So if you think of any more. We've got quite a few online questions. I'll also try and pull them where possible, Coby. The first one is actually from an e-mail from Stuart that we got. It's in regard to the pipeline. Would you mind talking a bit more about the pipeline as part of your presentation? Specifically, you have mentioned previously potential signings are being held as no one wants to go first. Is that a comment in relation to fabs, IDMs or product companies? And now that customers are signing, is that helping discussions with others?

So, first of all, human nature is such that we don't like change. And people like to -- what they're comfortable in to stay in their home, in their environment. So whenever change come, people resist it. And that's true for everyone, right, and in any domain. Specifically in this domain, it's true for the fabs, for the product companies. We had a situation where I can tell you with one of the fabs, I thought we were going to sign a long time ago. And then they were already convinced and wanted to go forward and then their product customer told them, hey, we're still not comfortable with this new stuff and whatever, and we want to do flash for another project and stuff. And suddenly, the whole thing comes to us screeching halt and they focus on flash and they come back to us almost a year later. So that's part of the challenge that we have in giving these guidelines and things. You never know how it will work even when the fab is already telling you, we're totally convinced we want to run with ReRAM. Suddenly, their product company tells them, "no, no, no, no. We really prefer flash first. Let's let someone else do that product. We'll do ReRAM the next one, right? So it happens. And -- but I think, again, and this is also true. The more people you see around you starting to use something, the more at ease you are with using it. And that's what's happening. onsemi really helped us. I mean it's -- there's before onsemi and after onsemi. And after onsemi, when people see such an important player in the market, that's signing up with this technology, they say, okay, they're talking. They're talking. Now again, some of them are dragging their feet and saying, oh, until we don't see DB HiTek with a product in mass production, we're just not going to move forward. We want to see it in mass production. And there are several like that. Others are saying, okay, we understand that if onsemi analyzed it and took that risk, we can feel more at ease and they still do evaluations and they still go through all of this stuff, but we're progressing. These negotiations, by the way, are just -- they take so long and their lawyers. I can tell you, let me go back to onsemi. If we go back more than a year ago, I was sure onsemi was going to close before September. It was literally ready for closure. And then suddenly, they tell me, oh, the lawyer is busy now. We're doing an M&A. I think they announced some acquisition in November or something like that. The lawyer just disappeared, and we were so close and the lawyer disappeared. And that's why it ended up closing in December. Now you asked me December 31, it could have been that the lawyer would have been busy two more weeks, and we couldn't have closed onsemi on time. So it's that kind of thing. It's -- we're doing our best on these things to push these forward.

Okay. There's a few questions on DB Hi-Tek. Two or three investors online have asked about an update on the quarter. You've already answered that, so we don't need to answer that again. But Stuart's got another question in terms of DB Hi-Tek. Are there still a lot of fabs, IDMs and/or product companies waiting for DB qualification? That is. Will that be the dam wall bursting when that's completed?

It's going to be definitely an important milestone. There are some companies that are telling us, we just want to see that happen. We just want to see that you actually close that qual. So I don't want to say it will open the wall, but it's going to be another crack in the wall. I think that the big thing will be some more major players signing up and we're pushing on that. I mean that's the big push, but it's definitely going to help us push forward.

All right. There's a few questions onsemi. Can you provide a little bit more detail on the final qualification with onsemi?

I guess onsemi, we can't really talk -- what we can say is we taped out, okay? Now this is now being manufactured, guess where, in onsemi, okay? And you can understand that they consider it a high priority. So now I don't control their priorities in the fab. I mean it takes many months to manufacture. They can give it higher priority. I don't know what other stuff is running through their fab line right now and where they're putting these wafers relative to products that they manufacture for their customers to actually get revenue. So, but they're manufacturing them, and we'll get them back. Once we get them back, as you know, we start doing all the testing and verifying and then seeing that everything is in a good shape. And hopefully, at that point, we can start qualification. The good news is because it's, again, it's the fab that is also the customer and so on, when we want to run more lots, if you remember, we need to do qualification on 3 separate lots that run independently. So, I mean, those kind of things, I imagine. Again, I'm not committing to anything, but I imagine they will probably give priority to get these lots through so that we can do the qualification and they can get their products out.

Some of the investors are pushing a little hard at it, Coby. I know you probably can't answer this, but given onsemi was talked out in early October, is there a chance you could be receiving royalties from onsemi in full year 2026 fiscal year?

I guess Unfortunately, again, they will -- you can assume that they'll want to have this in their products and they'll want to push this forward. But their schedules, we know some of them. By the way, we don't know a lot. They keep it secret from us as well. And even what we do know, they made it very clear that it's under NDA, and we shouldn't be talking to anyone about it.

Maybe to rephrase it as Warren does here, once qualification is finalized, how soon will you have products ready-for-sale potentially?

I really try -- I try to be as transparent as I can with you guys, but onsemi was very clear, don't talk about these things.

Maybe we can just say in general, it takes between 18 months to 24 months for.

Yes. From the day they start doing a design of a product, it's -- the rough schedule in general, again, talking generalities, it's I would say, between 18 and 24 or 30 months. It depends on the product and the complexity and whatever. But that's kind of the time scale to get product out to mass production.

Okay. Maybe a final one on onsemi before we go into other matters. The current tape-out with onsemi, is that the end design final product that's taped out in order to reduce the time?

No, no. No, it's the test chip. You need to remember, this is the first ever tape-out of ReRAM. So it's a test chip that we designed. I think I presented it a while back. It's a complete test chip. It has a processor. It has SRAM on it. It has a bus and everything. I mean it's a complete basically, it's a complete SoC, okay? We did a complete system on a chip, which has our ReRAM in it where you can really load programs in and run them. And so it's a real system that is going out, but it is our design for test purposes at this point.

Okay. We'll move on to other topics. So, from Chris, Coby, firstly, congratulations on a great quarter. How are things progressing with automotive companies? Will this be a slow burn in terms of them putting pen to paper?

You can answer that one also, right?

Well, I want to say automotive is definitely one of the key markets and a lot of interest in that domain. We get -- in modern vehicles today, you get thousands of chips in there and they are more advanced than what you might find in other IoT devices. There's the push over there, we feel it very clearly. I think onsemi is a very good application. onsemi is all about automotive applications. So I feel very confident with these prospects over there.

I would add just the thing about automotive is human lives depend on. And that's something you need to remember, the regulation that you have in different countries related to it, the standards that you need to meet. It's not just AEC-Q100 that you need to meet. There's ASIL and there's ISO 26262, and there's 1 million different qualifications, if you want to call it, that they need to pass and they need to show that people won't die because, right? So it's very, very critical automotive projects normally take much longer than an average SoC. And that's something that people need to understand. They will engage with us now, and they see the potential. And I mean, a lot of these automotives are actually analog. Remember that graph there, a lot like onsemi. So a lot of these automotive guys are analog and they will engage with us now. They can already see the AEC-Q100, they can already see the potential. But how long will it take until it's actually in mass production, automotive projects take longer.

Maybe just to add on that, it's also stickier.

Yes. I mean once they work with you, the cars -- the same model of car, and it's not just the same model of car. The subsystem is used in many future generations of that car. So once you're in, you're in for decades. I mean those are the type of projects that you sign up and you know royalty is now going to come in for a long time, you're going to have that royalty stream.

Okay. The next question from Jason is around the security product. Can we get a sense of what a security product means in terms of the use of a Weebit chip?

Well, secured products mean one of the advantages that ReRAM has is the security. The fact that the NVM is integrated with the main SoC means it's much more difficult to hack the products. You get secured products in payments, secure payment. You get this in your NFC products in your smartphone. Also secured products in automotive, right? It's extremely important to make sure that there's no way to hack into your autonomous vehicle or whatever it is. So these are the types of products that we're seeing over there. And the NVM is a critical part there, right? It's one of those.

The NVM also -- I mean, ReRAM has some unique characteristics about no two ReRAMs are identical. And so people use that, for example, for security keys or things like that. So it's used in many different security applications. But that variability that you have between ReRAMs and the fact that no two ReRAMs are the same is something that a lot of people like and use that in the security.

Okay. The next question is around the architectural agreement. So is the architectural agreement still alive? And if yes, can you comment on what nodes would be part of the architectural agreement?

So is it alive? Yes. One of those fabs that we were talking to is, again, an architectural agreement is also much more complex than a regular agreement, and that's one of the things that's causing it to drag forever. But yes, I never -- if you would have asked me in 2023, do I think I'll be sitting here at the end of '25 and still saying, yes, it's alive, and we haven't signed it yet. I would have said you're crazy. How long will it take? But unfortunately, it's definitely alive. It's definitely alive, and we are working with those guys. But once we announce it -- yes, I mean, once it's done, we'll announce it. I can't talk about it before that.

Maybe on GlobalFoundries, there' a question here. Maybe can you provide an update on if the GF chips were qualified?

The best of our knowledge. I mean, in their announcement, they didn't mention qualification. If it were qualified, I imagine they would want to say that. So they didn't mention it. They also didn't mention the word automotive anywhere in that announcement. So, two things that we didn't see in their announcement. Now they started much later. They are putting a lot of emphasis. I know that they are putting a lot of emphasis to push this fast through. But still, it's a lot of work, and we don't really know beyond that where they stand. But we're relying on their announcement. If I were them and it were qualified, I would say it's qualified. I want the customers.

Christian is pretty keen to know, are you working with any mobile phone product companies?

Also, our...

Among enough.

Among, among, yes, we're talking to so many different product companies, and there's also those.

Yes, smartphones, they have tens of different chips inside from the big SoC that runs the application processor that's usually much more advanced in terms of process node to a lot of power management devices that you have there and smart payments and all these things. So, I'm sure eventually, we'll find ourselves in one of those smartphones, yes.

Okay. Turning to SkyWater. Maybe just an update there on the current relationship status.

I think right now, it's pretty obvious they are totally focused on their R&D services, and we're a customer there. So, I mean, we definitely -- we're a customer right now of SkyWater. The fact that we're qualified there is great because now R&D can actually do a lot of testing there, and they are faster and cheaper than Leti. So a lot of things we're actually running through them to save money and to do things faster. In terms of working with them, to be honest, right now, they've disappointed me so much that even they acquired a new big fab in...

In Texas.

In Texas. I don't know what they're doing there. I don't see I don't want to bash whatever. But right now, there's no real discussion about that. And by the way, we're talking to such bigger fabs right now. To me, they are lower priority. I mean I'd rather close agreements with some of the big guys than with SkyWater anyway, so.

There's a couple of questions around discrete. So the first one is, do we risk our competition beating us to discrete ReRAM while we focus wholly on embedded?

So, first of all, we're not focused wholly on embedded. There is work in the background. It is low priority, admittedly, but there is work going on all the time. I mean just in September, I was at Leti with the R&D team, and we were talking about discrete and how we push this forward and looking at the different options and so on. So it's definitely not dead. It's definitely something that we know is important and we want to push forward. I personally believe we -- I am not aware of any other company that is working on discrete ReRAM right now. So I don't think we're risking someone beating us to it. I haven't seen anything anywhere about a company that's trying to do discrete ReRAM right now. So it's something that people would love to have, but it's a big challenge. I think that's -- again, going back to Weebit's differentiation, it I'm hesitating to say this, but I really can't think of another company in the world right now that has the combination of team, technology, resources, management focus that can actually do discrete ReRAM. I really can't think of that.

We had the Fujitsu that was doing some ReRAM for some time.

Yes, they were trying to do, but that was it. Yes. So there's -- I think we're definitely -- I don't think anyone is going to bypass us. It's true. It's not a high priority right now, but it's still working all the time in the background.

And given it's not high priority, the next question sort of alludes to this. So can you give an indication of the time line to discrete? And have you yet identified a preferred selector architecture?

I can't give any time line. It's still a lot of work. It's really a lot of work to do, and it's going to require R&D and we are exploring all kinds of talking about architectures. We're still at the phase of -- we're trying one and then we're thinking maybe there's another potential. And so there's different options that we're looking at. And yes, so it's still work in progress, and it will take time.

Stephen's got an interesting question here. So are there any restrictions on Chinese companies obtaining or manufacturing ReRAM?

The U.S. has a black list of companies that you can talk to. So those are definitely off of our target list. But beyond that, we don't have a real restriction. I can tell you that I'm just giving preference to anyone who's not Chinese right now because of the risk. I mean you don't know if a company that you're working with might end up on the black list later on or things like that. We've already seen situations where companies that we were considering working with that actually kind of ended up in the black list. So I'm very cautious. We have so much potential right now in the rest of the world. And China is a huge market, and you can't really ignore it. But at the same time, Weebit is just entering this huge vacuum. There's an unbelievable vacuum out there that we can fill. And there's no -- I mean, why take the risk of trying to work with a company where there might be an issue when you have so many others.

There's a couple of questions from Jason and Serena on EMASS. So maybe just to make the generic. Has the EMASS collab progressed? And can you maybe just talk to the talk that they're progressing towards a 16-nanometer form factor, which seems to be beyond Weebit's announced capabilities. Maybe just comment on that.

Yes. Well, we worked with EMASS, it was earlier in the year towards this demonstration that we have. It's a beautiful demonstration. I think recently, we also uploaded a video that shows this demo. I think it's very nice. It has great potential. They were using a 22-nanometer process node, and they have the option to continue and work with us. We're in discussions, but there's nothing much to actually report at this stage. So, let's see how this progresses.

Okay. We'll go to more market general questions. So anything happening on the AI front?

A lot.

This guy and his team are -- I think that's the biggest activity.

There's a lot going on, on AI. AI is divided into multiple steps. We're seeing a lot of interest in integrating our ReRAM into an AI SoC, just to get the memory closer to the processing elements and reduce the amount of data movements. And at the same time, we're seeing increased interest also in what's called in-memory compute -- this is where the computation is done within the memory element themselves. We're seeing a lot of interest from research institutes. I would say that for the past five, six years, maybe five years ago, this in-memory compute was just a concept that people were talking about. Nowadays, this is in real research, but not just within universities and research centers, but also large corporations and companies are investing heavily in this domain. I think it's still not ready for production, but we're getting there. We're getting there every year. It's getting closer and closer. I wouldn't put any time line on this, but it's there and the interest that we get from customers and from such partners is very high.

And I guess I would just add, Edge AI, I talked about it, I think, in the previous period time that I was here in Australia. ReRAM is really a natural fit for Edge AI, and that's a domain that's really growing rapidly and the demand there is growing. The challenge is a lot of these guys want to work at 22 and below. And right now, we actually don't have that to offer them. But some of them are working in larger geometries, and we are relevant. So, but in general, there are a lot of calls even from big name companies who say, hey, we saw you were working with ReRAM. You have ReRAM. We're looking at it for Edge AI and what do you think? So I mean, it's definitely something that a lot of people are looking at ReRAM for Edge AI.

Speaking of Edge AI, I've got a question here from Jason on Edge AI. So can you maybe just clarify, has this been signed as a product company or in negotiations?

So, unfortunately, we gave all the information we could on the product companies in the quarterly. Companies are still trying to be more -- I mean, eventually, they will be announcing their products. And I think these things will become public. Hopefully, they'll let us start talking about it more. I guess, beyond the fact that right now, the three product companies, what we mentioned are U.S.-based, I can't really talk about which application -- I think we -- for the first one, we mentioned they were a security company. The others right now we need to be really careful with them. So that's what we can say.

Maybe touching on Samsung. So what's Samsung's focus these days with regarding NVM? Are they MRAM focused?

So, Samsung, I mean, what they have in mass production right now is MRAM. And what they're offering their customers is MRAM, I guess.

I think it was public that they had some ReRAM.

They had a big ReRAM project, which they ended up shutting down because it just -- they couldn't get it off the -- working. Beyond that, I don't think we can comment.

Michael's got a technical question here. How are you going in terms of reducing the process node to 10 nanometers?

Well, I said we're talking to a lot of fabs. And obviously, that's the direction we want to go, and we want to get fabs that work in the smaller nodes. So, eventually, we'll be there, right? It's a matter of pushing. And you can see -- again, that's part of that "perceived risk thing that in the beginning for the companies, the smaller the geometry, the higher the perceived risk. And so that's why you've seen that we started with 130. Now we're at 65 and then we're trying to push down, and we'll be getting there.

Okay. The next question is around your technology. So once it's in a customer product, will the performance results help sell that Weebit technology to other possible customers?

Definitely.

I mean the ReRAM has big advantages over flash and people will see it and people will see the competitive advantage of the products that have ReRAM in them. I'm sure that, that will help us.

Yes. And when are you looking at reducing the smaller nodes?

Well, as soon as these guys will sign that damn paper.

But we're confident there's no technical barrier to a 1x geometry, knock-knock.

At the 1x, definitely not. Below that, there's more work to be done. But at the 1x, we've already done quite a few -- yes, I mean, I'll even say we have PDKs. We did simulations. We've seen that there's no issue.

Is ReRAM a good fit with the new generations of humanoid robots under development?

Definitely. Definitely.

Definitely. For multiple functions.

Yes, for exactly. I'm thinking for a lot of different functions there. We talked about AI, which is part of it. It's really for many different functions.

AI, it's the motor control that you need to run, which is it's a power management. There's many different functions within this.

Maybe a question on one of your competitors. What are the key differentiators between Weebit's ReRAM and TSMCs?

From what we know, and again, we don't know enough about TSMC from what we know. I would say, first of all, it's the company focus. It's not the actual technology. We -- I don't think we can really comment on speed or performance or power or whatever. We don't know enough, and I don't want to go into that. But I think the real difference is the company focus. Weebit has the design team that is there to help -- and we know because we had customers who actually called us up and said, we talked to TSMC, we wanted to use their ReRAM, but we wanted them to do some modifications for us, and they wouldn't do it. They said, God help, basically. TSMC is a fab. They want to manufacture. They don't want to start dealing with modifying. I mean this whole ReRAM is just an enabler for them to sell wafers. So they have several versions. It's a good ReRAM. First of all, let's make it clear. TSMC is TSMC, and they have a good ReRAM and people are using it. And I'm not going to say it's anything bad about the ReRAM. But Weebit, the advantage is that we have everything around, and we will work with the customers to give them what they need to tailor it to give them a better solution. I believe that the investment that we're making in R&D, and we are working on new concepts in manufacturing ReRAM and on improving the whole manufacturing process. I don't know how much TSMC is continually investing in it. Again, it's not their core competence. It's not something that they plan to make a lot of money off of. I mean it's -- right now, they are making a lot of money off of it because they're the only ones who have ReRAM. So they put this big margin on there. We tell people ReRAM doesn't add more than whatever, 7%, 8% to the cost of a silicon wafer, they say, but TSMC is asking for 40% like, well, because they can, right? So, but in reality, they're not -- this isn't what they're going to make money off of. And at a certain point, there's a limit to how much they'll invest in this R&D. So I believe that over the years, we will continually improve the technology. And at a certain point, and it's not going to be immediate, but at a certain point, I believe we have a good chance of having a better ReRAM that some customers will want and will push TSMC to license the ReRAM from us.

We're on the last four questions online, and then I'll throw back to the room for any final questions. Very general questions. So are we going to see any takeover offers in the near future?

I guess everyone knows what my answer to that.

And on the topic of cap raises, any cap raises in the future?

With $91 million in the bank, I'm focused on getting the deals done. I really don't think I need to think of cap raises.

There's maybe more of a comment here rather than a question in terms of maybe getting other directors like Atiq to join these webinars maybe once a year. It's good having Eran on, maybe introducing other directors as well maybe.

Honestly, I never thought -- I mean, we have Dadi coming every year to the AGM, and he's there. The others are nonexecutives, so I normally don't think about getting them involved, but that's -- let me think about.

A couple of more have come in at the last minute. Outside of the foundry deals and partnerships you've already talked about, is there any other area, potentially a new market or application that you think could quietly become a big growth driver for Weebit over the next few years, something that's not really on the radar at the moment?

We look around. I mean that's this guy's job, right? Business development is his title, right? And we're always looking around at how -- first of all, how ReRAM can be used in different places. And then every once in a while, there are ideas of complementary things. I mean this is the type of thing that I think any company is always looking at, at what's happening in the market happening around. And I mean, when things happen, they happen, right?

I think one interesting market that we're seeing that we get a lot of questions about data centers, AI, but data center side, right? So very focused on Edge AI, and we're doing a lot of stuff there, and I hope soon we'll be able to talk more. But on the data center front, what we're seeing is not necessarily the NVIDIA chips, which run at a much more advanced process node, but the power management, which is a critical issue when it comes to data centers, right? The amount of power and the air conditioning that runs in there and all that. So power management is critical. And one of the markets, for example, for semi is data centers, power management for data centers. So this is definitely an interesting way into those data centers for resistance.

Okay. We've probably only got time for two more questions, maybe one from the room.

All right.

Okay. So the third last question. From your perspective, how should shareholders best interpret the information that you shared? And what are some of those indicators that you can suggest that investors should focus on to understand your future momentum?

I think Yes. I -- honestly, it's hard to talk about yourself and whatever. But I think Weebit is a very conservative company, which is just focused on doing the job. We basically try -- I think you can see -- first of all, I try to be as transparent as I can. Second, you can see the targets and how we progress from year-to-year and slowly. It's -- actually, in semiconductor terms, we're really moving fast. So I know it looks like forever and -- but we are moving, and it's really steady as she goes, right? So I think that more than anything, I hope that shareholders are looking at how things evolved and are seeing that we went -- I remember when I joined the company, we had just the first bit cell and then the array and then we got to the full chip and we qualified and we went to SkyWater and we went to DBH and we got to onsemi. And so I hope you can see that trajectory. Now you can also see the cash. I mean, fiscal year '24 was $1 million revenue, fiscal '25 was $4.4 million. Now just this quarter, you can see already how much cash is coming in. So you can start understanding where it's heading. I really just hope that people see that we're serious. We're doing the work. I don't like to blow things out of proportion. I don't like to go -- when things don't work well, we don't go into, oh my God, everything is falling apart. We're -- it's a very experienced team. We've been through endless hurdles in our lives. And at Weebit, believe me, we've gone through quite a few. And we just -- we hit the hurdle, okay, what do we need to do? We fix it, we move on, and we continue to progress. So I just think people need to look at that and see how every quarter things advance.

And this is a good one to finish off from the online questions at least, 10 years, with your crystal ball, where do you see the company in terms of scale or size versus your peers?

Or the market.

Yes, I'm not giving any guidance because this -- in Hebrew, we say prophecy was given to the falls. I mean that's what's written in the Bible, prophecy was given to the falls. So -- and I don't consider myself such of it. So -- but I mean, obviously, I believe Weebit will be growing. My goal. I can say what my goal is. My goal is to be the leader, the #1 ReRAM company. And I think in 10 years, it's going to be not just ReRAM. We will expand beyond ReRAM. There is a lot more out there. I mean, who knows -- again, people asked about M&A. Things happen when they happen and when you have opportunities. But over 10 years, who knows, you can assume that there might be some opportunities. the whole market, AI, we don't have a clue where it's going to head. ReRAM is a natural solution for AI. So I believe you're going to see a big AI market and you're going to see Weebit as a big player there. That's, again, my belief and my goal, what I would like to do, no guidance or any commitments or whatever.

Do we have a final question in the room?

Just regarding the test chip that's been manufactured and in some cases, qualified. Is there any commercial viability to sell that as it is since it's already gone through the full process?

It's a very simple system on a chip. I mean some people might want to actually use it. I don't know.

There is a possibility, I think. For certain applications, some IoT applications, it could be a good fit. But I definitely believe that for a true product, you would need at least to add some kind of interfaces and real-world interfaces into the chip or...

I thought that actually was the idea that was made quite generic and versatile for people to test out.

Yes. It's very generic and versatile. I agree.

But for commercial purposes, they'll probably want to tag on some more. And to be honest, I haven't really thought much about it. But if a customer comes and say, "Hey, your test chip is really interesting. I just want to tag on some things. We have no problem to license the full test chip.

A quick question about patents. Could you tell me or tell us roughly how many patents you would have covering ReRAM compared to your main competitors like TSMC?

So we're not following up on specific patents from competitors. I think...

Just numbers.

Yes. So I think in the last quarter, we announced six new patent applications were made and we were at more than 90 patents for us. With regards to competition, I don't have the answer.

Yes, we actually -- it's -- in the industry, people try to avoid researching the patents of others because you can only get into trouble. So...

We took get someone else to do it for you.

I guess it was less interesting for us. As long as we know that to the best of our knowledge, we're not infringing on anything, and that's what we care about. That's what's important.

Unfortunately, we've run out of time. So that concludes the Q&A session. I'll now hand back to Coby for any closing remarks.

Well, I guess, first of all, I'm always glad to meet you guys and the people that are out there. as I said before, Weebit is moving forward, progressing, making good progress. You can see it. We definitely plan to close more deals this year. We -- that's the big focus right now next year to already get into that tornado and really get this thing moving. We have an amazing team, and we have amazing shareholders that stick with us in the ups and downs and all. And yes, I mean, the numbers, again, what happened this quarter, I'm repeating, don't expect every quarter to be like this and going up, but it's good, and we are going to continue to get money from customers and the numbers will grow over time. So, it's good, and we're moving forward.

Okay. Thank you, Coby. Thank you, Eran. Thank you to all the people who attended in person, and thank you to all those online who can now disconnect. Thank you.

Thank you everyone.

Thank you.