Weebit Nano Limited (WBT) Earnings Call Transcript & Summary

Good afternoon, everybody, and thank you for joining us this afternoon. My name is Eric Kuret from Automic Group. I work with Coby and Weebit Nano on their Investor Relations. Delighted to have Weebit Nano's CEO, Coby Hanoch, here with us this afternoon. The company released its Q4 FY '24 quarterly activities report yesterday, as well as the DB HiTek tape-out and also an update on the Board. So a bit of news flow all fell yesterday. And so delighted to have Coby here. [Operator Instructions] We do want this to be a very conversational update. So Coby, I will hand over to you. Obviously, there was a bit of news flow yesterday. Perhaps to those maybe not in the tent had thought maybe the company has been a little bit quiet for a while, so maybe give them a bit of an update through the quarter and what's been happening at Weebit.

So hi, Eric, and hi, everyone. Yes, I know. Yes, I've been getting feedback from people who are saying, "What's going on. You guys are a sleep or something." And it's really a very different situation. Actually, there is so much happening right now at Weebit, and we are working around the clock. I think, first of all, let's just take a quick look at what we've done over the past year. We just finished the fiscal year. And a year ago, we were still basically just qualifying at 85 degrees and we had just SkyWater as our first foundry that we were working with. We were trying to really push forward with many others. The market itself was actually still getting to understand that ReRAM is here, but we were still pushing forward. I think over the past year, well, obviously, we signed DB HiTek, which was a very important step for us. Yes, this week, we taped out, and that was another major step forward to qualifying with DB HiTek. We expect to do that roughly around Q2 of next year. So before the middle of next year, we should be qualified there already, and now we're working with them and our customers. There has been a lot of work on qualification for 125 degrees. We actually demonstrated we can achieve 150, which is something that, in the ReRAM space, actually, in general, in NVM, is a very tough task to achieve. And we're now even working on qualification for 150 degrees and high endurance going to 100,000 cycles already. We demonstrated that. We have the wafers that we manufacture the GF 22. And I think that was another important step forward. The market reaction has been really very strong for that demonstration, showing that we can go below the 28 mark where really flash gets stuck at 40 or 28. So there's been a lot of technical progress. There's been a lot of business progress. Well, obviously, we brought onboard Issachar who's really giving a strong push. And today, the level of engagement that we have with so many of these big foundries, these big IDMs, is just crazy. I mean the team is really being torn with so much activity. And then some of those are already at the negotiation phase. So we actually got to the point where with some of these foundries and IDMs, we are negotiating the business terms. And again, this is looking good and moving forward. Overall, I'm feeling extremely bullish. I guess you saw it also in my, Dadi and Atiq's action in June that we demonstrated our belief in the company, and I'm really looking forward for the next year.

And just to clarify, by demonstrating your belief, you referred to buying shares on market for those that weren't aware. Thank you, Coby. It definitely does sound like there is a lot obviously happening there, so thank you for that update. Maybe before we dive into more specifics around what Weebit's doing on the various streams and where certain things are at, why don't we start with sort of the broader market? How are you seeing the broader memory market right now and specifically, ReRAM's role in it? If you can give us an update there.

Yes. It's really great to see how the market is being transformed. People have been looking for a replacement for ReRAM for so long. And a few years ago, when MRAM came out, everyone thought that's...

Coby, just to clarify, a replacement for flash, right, not the ReRAM?

For flash, I'm sorry. For flash, I apologize. So everyone has been looking for a replacement for flash for so long. And a few years ago, MRAM appeared. People believed that, that was going to be the solution. And the market, for quite some time, was really trying to push it. Now we talked many times about the limitations of MRAM and the fact that it's expensive and the fact that it's sensitive to magnetic fields, et cetera, et cetera. But even as late as a year ago, if you would have looked at TSMC's road map, for example, it had MRAM as the main thing and ReRAM was a very negligible thing on the side there. Today, and just a few weeks ago, TSMC released their new road map. ReRAM is all over it. I think there's less than 10% of it that has MRAM. There's only 1 line that's kind of half ReRAM, half MRAM and all the other lines are just pure ReRAM. TSMC has gone through the realization that ReRAM is it. Now they're selling ReRAM. They have customers for it. They see the results. They see, on the other hand, how customers are shying away from MRAM because of all of the issues. And they are a strong trendsetter. And when TSMC makes such a decision and such a change in focus, it really reflects what's happening in the market. Now beyond TSMC, you have Infineon. I think people know, because that's already a little bit even earlier, Infineon announced that their future MCUs for automotive, they're all going to be based on ReRAM. And that's another important thing. We've seen very big and important consumer companies who basically made it clear they will never use MRAM because of the fact that their products go everywhere, and there are so many magnetic fields in so many places. They have a high risk of having the memory not work if it's MRAM. So I think that what we've really seen over the last year, and it's really recent, we've been saying that the next NVM is here. And we've been talking about and we've been seeing that people accept ReRAM. But now it's really you're seeing that flow in the market. Now it's really moving towards ReRAM. Today, I don't think anyone has a doubt that ReRAM is the solution, that will be the main NVM solution moving forward. And I think that's another very big thing that's been happening in the market over the past year.

Great. I mean that sounds super positive. So I guess then the logical question is, why hasn't Weebit closed any deals yet? What's taking so long?

Yes. That's the big question I've been asked for so many times. And it's kind of interesting. There are industries where things inherently, they just take time. I think in Australia, everyone knows the mining industry. And for them, the fact that a mining company takes a few years to go and search and survey the grounds and start digging and until they actually have a mine operating there, it can take many years. With pharma companies, people know that it also takes time to go through all of the FDA regulations and so on. Semiconductors is pretty much very similar in the sense that it's not a simple technology. It's not a simple market. Developing a technology like ReRAM, we need to remember, today, there are only 3 qualified ReRAM solutions in the market: TSMC, UMC and Weebit. TSMC and UMC, the technologies that they're using have been under development for more than 20 years, 25 years. Even Weebit, we are around 10 years, but we're leveraging a lot of know-how and experience from Leti that's been researching ReRAM for more than, I think, it's also about 25 years. These are technologies that take a long time to develop. And then when you engage with customers, again, these fabs are billion-dollar facilities. Everyone is cautious. Everyone wants to make sure that we won't contaminate anything in any way, that our process is fit for them. And we talk about transferring the recipe. It really is so much like teaching the foundries how to bake a cake, and that's why we use the term recipe. Now try to think of it, maybe to make it easier for people to understand, we have the recipe for this really great cake. And part one step of it is, for example, whipping the cream. And we have our mixer that we use. But the foundry has a different mixer, which is slightly faster, negligibly but slightly faster, which means that it will reach the thickness of the cream maybe a little bit earlier. And you need to start doing these adjustments of, okay, so you have a different mixer and how long do we need to use it or how do we use it, et cetera, et cetera, I mean it's those things, but we're talking about such delicate things that it takes time to do it to measure the manufacturing. I think people already learned, just 1 round of manufacturing takes half a year. And when they come to us and they say, "Oh, we want you to demonstrate this. We want you to demonstrate that. We want to do this." The simulations, and sometimes we need to do these short loops or things, these things just take time. Now these foundries are also very concerned. Their customers have specific needs. Sometimes it's voltage. Sometimes it's power. Sometimes it's leakage. Sometimes it's cell sites. And we need to demonstrate that we can achieve those things. So these whole procedures takes time and then, obviously, were not the only thing that the foundry is doing. Sometimes they just have a crisis on another line or they have to bring up something or whatever, and they suddenly halt the discussions for half a year. So we're making very good progress. We're constantly moving forward. I think I can say with confidence, I'm not seeing anyone who's saying, "Oh, we don't want this," right? We have different situations, but we're moving forward, as I said. With some of them, we already got to the point where now we're discussing the commercial terms. So before the end of this year, I still believe we will be able to start getting some agreements, at least 1, hopefully, even 2 agreements. And even product companies, we're making now progress with them and looking for the ones that can manufacture SkyWater with all of the limitations and so on. So yes, we will be moving forward. We will be announcing agreements. And I think once we start announcing the first ones, people will also be less concerned, right? Now nobody saw a working silicon with Weebit in mass production. As we move forward, I think the resistance level is going to fall.

Great. Thanks, Coby. There was a question that came through saying how much longer until you sign a deal. So I think you covered that in your response then and your expectations. So Coby, you talked a lot about the foundries and you also mentioned product companies. Thinking about product companies, are you seeing any specific segments which you think are more open to ReRAM?

Yes. Everyone needs ReRAM. I always say that any electronic device needs ReRAM to store the code and data. But there are those that really need it now. One of the domains that's growing rapidly now is power management. We used to just plug things into the wall. Now most devices have smart chargers. Anything that charges with a USB-C, there's a whole protocol that goes on during the charging. And I think some people noticed or realized it starts charging slowly, then it charges faster, then it slows down. Suddenly, these power management guys, for the first time, they need to have algorithms. They need to have logic running in their devices. So they need to add an MCU, a microcontrolling unit. And MCU has to have a nonvolatile memory. So suddenly, these guys, for the first time, they need these things. Now their circuitry is very delicate. It's an analog circuitry. And flash is a front end of line technology. If they put flash next to their design, it impacts their design. There are a lot of considerations that they need to take into account, and it's not easy. ReRAM, in this sense, being back end of line, coming on top of the design, the design doesn't really care about it, they can do the most optimal design. For them, this is really a great solution. And when we talk about BCD, one of the key users of BCD, like we are doing with DB HiTek, is for PMIC. So for those guys, this is a key consideration. We have the automotive guys. Today, again, the number of chips in a car is growing like crazy. It's more than 1,000. Even people say already it's 1,500. I remember in meetings with shareholders just 3, 4 years ago, I would say it's a few hundred. So it's growing rapidly. In automotive, some of the chips need to operate at high temperature. This is a critical element for them. We've demonstrated that we can operate at 150 degrees, which is a very important milestone for them. And so this opens the door. And today, we are engaged with some major, some of the leading, automotive companies who are looking at our technology. Again, the cost and the simplicity is important for them. They've been trying to use MRAM. They've been having a lot of issues with MRAM. So this is really a key thing for them. There's edge AI. One of the things that's been happening in the market is people thought that everything will be done in the big huge servers on the cloud. There would be just edge devices that would collect data and just send it to the cloud. Now it's suddenly clear that the amount of data that needs to be sent to the cloud is huge and the bandwidth and the power requirements and everything is just crazy. So now everyone is realizing that the edge devices need to be smarter, and they need to do what's called inference. In AI, you have the learning phase where you learn and collect a lot of data. You analyze it, you learn. And then there's the inference where you use what you learned in order to recognize things and in order to reach conclusions. The inference part is you're basically taking a lot of weight and a lot of coefficients and you're reading them all the time, which is perfect for a nonvolatile memory. And since a lot of these devices are now at 22 and going down to the teens, and even below in the geometry, flash is not an option for them. Well, if they want to use flash, they need to have a separate chip. And then it's a 2-chip solution. It's costly. And there are a lot of side effects. Every time you turn the machine on, you need to load the coefficients from the flash into the chip, into the SRAM and the chip. When these things operate on batteries, you want them to sleep as much as possible. You want to turn off the power as much as possible. SRAM, you can't turn off. So you either turn off and every time you need to reload from flash, you don't really go to sleep and you're wasting a lot of power. So having embedded ReRAM in these kinds of devices is really a very, very optimal solution not to mention the fact that the ReRAM bits are smaller than SRAMs, so you can actually have even more coefficients there. So this is a trend that's now picking up. And I can go into other markets, but I think these are really leading in key markets. And just one last comment, I talked about the temperature. Even markets that don't require work at high temperature, the mere fact that we do manage to work at high temperature shows them the robustness of the technology and gives them confidence. And that's really one of the things that's been helping us. When we talk to these product companies, they realize that this is a strong and robust technology that they can actually have confidence in.

Got it. Thanks, Coby. Just when we've been talking about product companies, the questions come through from an attendee just asking if you can give us an overview of the typical time line for how a product company would develop a new product and how the components of semiconductors would go into it.

So a product company, and we're talking about the ones that are doing systems on a chip, which are the leading ones, and today, I think I can say the majority or a very large number. These guys, basically, they need to make decisions on what goes into their SoC in the beginning. So they have the architects who are defining, "Okay, this is the functionality that we need," and, "We need to have a processor," "We need to have, whatever, a Bluetooth or something," "We need to have a digital processing unit," et cetera, et cetera, and "We need to have nonvolatile memory." The next step is they need to decide where they're going to get all of these components from. They're not going to design everything themselves. So they look at the market and they see, "Okay, we're going to take the processor from ARM, the WiFi from Nordic," whatever. And hopefully, they decide to take the nonvolatile memory from Weebit. So now when they make those decisions, now they need to go ahead and design their product. Designing the product, depending on the product, it can be 9 months to 1.5 years. I think the average we can say is about a year. Once they finish the design, they need to go and tape out their prototype, get the silicon back, test it, verify it and eventually do a qualification for their product to see that it's ready for mass production. This whole process, the design is a year and then manufacturing and testing and qualifying and all of that, again, that's something like 1.5 years, 2 years to ramp up until they get to mass production. Sometimes it's even more. Now they like to start it, I mean in order for them to choose Weebit or ReRAM as the nonvolatile memory that goes into this design, they basically want to know that it's qualified. And that's for them the stamp of approval, "This is ready for mass production." What we are doing now is we are approaching customers of DB HiTek and trying to get at least a few of them to realize that, "Hey, this is a robust technology. Yes, we're not qualified yet. It's still going to take us some months. But we're going to be there. We're going to do it. You guys have a whole year of design work to do. You can start it now and be the leaders in your segment." So it's not easy to convince them to do it for them. It's a risk, engaging before we're qualified. But we are pushing in that direction. We want to get some people signed up. And so the process is long, but the thing that you need to understand, once the company starts doing it, these products are in the market, some of them are for many, many, many years. And the royalties are going to flow over many years. And even once they put you in one product, then they start putting you in additional products, et cetera, et cetera.

Got it. Coby, so in the quarterly, I think the words we mentioned around customer engagements was like over a dozen. A question from one of the attendees, Daniel, has said, given that level of engagement, is there a limit to how many customers or foundries we could sign in the short term? Given, are there any sort of internal capacity limitations?

Well, yes, obviously, there's a lot of work involved in each one of these negotiations. And that's why I'm really saying the team is stretched so thin because we have these. And I, as a CEO, need to basically balance things out. Yes, I mean I'd love to just hire more people and work with even more companies because there is demand. But on the other hand, we have a certain amount of money, I'm feeling very good about what we have in the bank, but I don't want to waste it too fast and lose it too fast. I think the shareholders are expecting me to work in a controlled manner to balance things out. We're engaged in a lot of evaluations right now with foundries, with IDMs, with product companies. And yes, it is stretching the team thin. As soon as we bring in an additional deal, 1, 2 agreements, we see that things are moving forward, we will be hiring again in a very controlled and conscious manner to enable us to work with more companies. So yes, we're growing. And we're growing nicely, I think, in a very good way. We have a lot of engagements. We're going to be signing some agreements before the end of this year. And we will be growing. But I think that's one of the things that hopefully, the shareholders are appreciating, that we're not just going wild and spending all the money too fast.

Got it. Thanks, Coby. A couple of questions coming through just on SkyWater. Sort of summarizing those comments is we had hopes that we would have had SkyWater customers using ReRAM's technology by now. So where are things standing up with SkyWater? What is the current relationship? And are you happy with the work that has been done? And how do you see the future with SkyWater?

Yes. So SkyWater is an interesting an interesting company. All of the foundries that I know basically focus on being foundries. DB HiTek, the only thing that they do is they manufacture wafers for their customers. And that's the big focus and so on. In the case of SkyWater, they actually have two parallel activities. They have R&D services that they give and they are a foundry. Now as I think you know, we've been very disappointed that they decided, they made a strategic decision, to focus on their R&D services. And you can see it in their quarterly reports, the R&D services are growing. The foundry services are actually going down. For us, that was a major blow. I was sure that we were going to have customers product companies, before the end of last year, manufacturing at SkyWater. That was the plan. I think everyone knows it, and it was a big disappointment for us. Now the first thing is the R&D services, we're benefiting from that. Actually, we're benefiting a lot from that. Now that we're qualified there, we are doing a lot of the, I would say, the D part, the development part at SkyWater. We're manufacturing a lot of wafers there. It's faster, it's cheaper for us. So we're doing a lot of the research at Leti. And then the actual running wafers, we're doing a lot of it at SkyWater. And that's working great, and we have a great relationship with SkyWater. On the foundry side, yes, it was very disappointing. It was disappointing to have product companies who actually were ready to sign agreements with us and move forward and basically, they got stuck because some things were missing. We haven't given up. And we are looking for those companies that are able to work with SkyWater despite the fact that they are missing some components that are important. And we're finding, we're finding some of these. So there are engagements now with some of these guys. It still isn't easy for them to work with SkyWater. That's why it's taking longer, and they need to actually see how they overcome the obstacles. But they do want to go and work with us there. And so I hope that we will be able to announce those customers, such progress, over there. And another thing that we announced recently is the relationship with Efabless, again, another way for us to bypass some of these limitations with SkyWater. We basically have signed this agreement. Efabless is kind of a service company that can build these prototypes for customers. Normally, it can be research institutes. Sometimes it can be big companies who just want to test things out, and they go to Efabless. And Efabless manufactures small quantities for them at SkyWater. Of course, if some of these companies will decide that they want to go into mass production, then they will have to have an agreement with Weebit and they already have the design from Efabless, and we'll be able to move forward. So we're hoping to see some business come out of the relationship with Efabless.

Got it. So I guess the logical next question is, sort of given that response on the foundry side with SkyWater, I guess what gives you confidence with DB HiTek and that you don't have the same situation?

Well, as I mentioned, DB HiTek, the only thing they do is being a foundry, and that's what they do. That's their focus. We can see it. I mean they're one of the top 10. They have big customers behind them, Mitsubishi and Sony and Toshiba, and even companies like Intel and Qualcomm manufacture there. So they are totally focused on being a foundry. And the only reason they engaged with us is because they saw that they have customers who can benefit. I mentioned, we're talking about the 130 BCD. They have these customers that are doing power management and other things which, until recently, didn't really need NVM, but now they do need NVM. And so DB HiTek doesn't have embedded flash in the 130 process, they need to offer an NVM, and they were, I guess I can say, smart enough, that's how I look at it, to realize that the world is going towards ReRAM and it would not make any sense to invest in flash, especially with the limitations of flash. And they've put what they've got on Weebit. So Weebit is going to be the nonvolatile solution for DB HiTek at 130. And the relationship with them is really good, very professional. These guys are really young. And they're cautious. And constantly, they want to make sure that everything is working properly and so on. Sometimes we just want to run forward already and they say no. Step by step and make sure everything is good. We did the tape-out now. I think everyone is already learning. That's a very important milestone. We're going to be waiting for the wafers to come back and do the testing and do the qualification. But I'm feeling very good. The simulations and everything are looking good. So I'm extremely bullish about DB HiTek, and it is very different from what we have with SkyWater. I mean DB HiTek is already an operating fab that has all of these components that are missing at SkyWater. I mean it's not a question of will they have these components, they have them now, and that's the difference.

Excellent. So just on the topic of the tape-out and then going to qualification. The question from Andrew, he said, are there any steps along the way between now, the tape-out and gaining qualification with DB HiTek that shareholders should be looking out for? Or is it effectively you just get to qualification?

Well, there's always the steps. We wait to get the first wafers. I think people already know, to do qualification, you need to have several lots that come out. So the day that we get the first wafer out, we can start testing if that's going to be an important day for us. And then we do the testing. We'll reach the point where we'll say, "Okay, these chips are looking good. We're managing to form them, to read, to write, to do the basic functionality, let's go ahead and do the qualification" and, of course, that's an important decision milestone; and then the qualification. So all of this, we have these steps along the way, and people have seen them already previously with SkyWater and even with GF, so we're working on it.

Got it. Thank you. Lots of questions coming through about 22 nanometers and GlobalFoundries. So can you give any update that you can provide for us just around sort of the GlobalFoundries’ relationship and what's happening on that 22-nanometer front?

Well, unfortunately, everything that's going on with GF, GlobalFoundries’, is under NDA. I think what you know is, on the one hand, GF has another ReRAM. They have their own ReRAM, which they are testing. That's public information. On the other hand, we've manufactured there, and we have our ReRAM running on their wafers. We're demonstrating them. We're getting very good results. I can't go into the details of what's going on between these 2 ReRAMs and the discussions we have with GF. There are many customers at GF that want ReRAM, that need ReRAM. We're engaged with quite a few of those. But it's work in progress, and I can't comment any more, unfortunately.

No problem. Thanks, Coby. There's a number of questions coming through just around you previously talked about a potential agreement or negotiation you're having with a foundry that wanted an architectural contract rather than a contract for a single node. So just there's a few questions there, saying how is that progressing? And is that still alive?

Definitely. It's still alive. Again, these are negotiations. I can't really go into details of these ongoing negotiations, especially since you'll probably realize many of these companies actually watch this video as well. And I need to be very cautious about how I talk about things. So we are making progress also with these guys. And we'll be announcing things when we actually get things done.

Got it. Just thinking and talking a little bit around sort of geopolitical things, obviously, semiconductor has been a hot topic. So are you still engaging with Chinese foundries? And if so, are these negotiations impacting your relationships with other foundries and IDMs?

Yes, it's difficult. This geopolitical situation is a challenge. There are, of course, companies that are on the blacklist that we clearly don't even get anywhere near. We can't even talk to them. I mean, the French government, the relationship we have with Leti and so on wouldn't really want that, and the U.S. government, et cetera. However, there are quite a few foundries and companies in China, which are not blacklisted. And you need to understand, even today, with all of this geopolitical situation, Intel, Qualcomm and many other companies still have a significant part of their revenues coming from China. They're selling. They're working in China. It's not like American companies are not operating in China. It's far from it. There's a lot of activity going on. So Weebit as well. We have a lot of work going on with the Western world. And obviously, I give some priority or advantage to those discussions. But having said that, we are not totally disconnected from China, and there are opportunities for us in China that we're pursuing. And it's a huge market. We can't totally ignore it. And just like the other big companies that are operating there, we are doing it with caution and realizing the limitations.

Got it. Thank you. Just a quick question that's come through, just a clarification around revenues and how they work with Weebit's model. The question is, could you confirm, with all agreements, licensing will come first then some product development, then eventually product royalties? Is that right?

In a high level, that's true. The normal way is that when you sign an agreement, you get a license fee. Then you need to put in all of the engineering work and you get paid what's called the NRE, nonrecurring engineering. And then eventually, when the products are in mass production, you get royalties. In the beginning, and I think I mentioned this in previous times, companies are feeling that there's still a high risk level with Weebit, so they are not willing to pay the license fee upfront. And they basically have it split it based on milestones. So as we achieve milestones, as we move forward, they will pay us and we will recognize revenue and all of that. So that's going to be happening. That's why I tell people, our revenue ramp, there's going to be a revenue ramp. It's going to be gradual because a lot of these payments are going to be done over time and not as a lump sum in the beginning, at least not in the first agreements. I mean it's natural, I understand the customer concerns, and we're agreeing to do based on milestones with the first ones. Yes, it's true, that's the basic concept.

Got it. Still a bunch more questions coming through. Just as I sort of thought through and group those, just a question, obviously, you announced yesterday, that Yoav would be stepping away from his Executive Director role, just transitioning to a Non-Executive Director. Maybe just give a bit of context around that.

Well, when I joined Weebit, it's almost 7 years ago, believe it or not, we only had 2 engineers at the time in the company. And I really needed an expert in the field to help me understand and just make the decisions. And so Yoav, I remind people, he has a PhD in nonvolatile memories. He was the Founder and the first CEO of Tower, which is one of the top 10 foundries. So he has a lot of knowledge. And we basically made him executive so that he could help me in those early days. And his help was unbelievable. I mean, he really, really helped me push things forward, and he went with me 2 meetings with the different foundries and so on and so forth. However, over time, Weebit has grown. Weebit has become stronger. We hired a very strong management team. By the way, several of them used to report to Yoav. Ishai, our CTO, who's driving all of the device and process work basically was the Co-founder of Tower. He worked under Yoav at Tower for many years. Ilan, who is driving all of our analog and digital design work, also worked at Tower under Yoav. So now that we have these very experienced VPs, the need for Yoav has gone down. So I really thank him for all of his support. But I think at this point, it makes sense for us that he will become a Non-Executive Director. He's definitely staying on the Board. I'm still going to be consulting with him just like I do with Dadi and with Atiq. And by the way, now that we've become a larger company and analysts are looking at us, et cetera, the analysts want to see that the Board will be more independent. So it all kind of worked together. I think it was the right time to make this decision. It's a good opportunity to thank you all for all of your support over the years. And Weebit is now strong and moving forward already. This baby is growing.

Got it. And there is a further question just around the Board composition and, I guess, how each individual is contributing. How are you feeling about the Board of Weebit?

Oh, it's an amazing Board. I keep getting these comments from other CEOs and from other people, how amazed they are at the Board that we have. I mean, first of all, even big, large companies don't have -- I can't even think of someone who has the composition of Dadi, Atiq and Yoav on the same Board. Having just one of them is oh my god, having the 3 of them is unbelievable, and they work so well as a team. They are just so amazing at brainstorming together and giving us basic guidance from their experience, from their understanding. It's just awesome. And then we have Ash, Ashley, who's been with us from the beginning. He has a very good understanding of what shareholders in Australia think, how the market reacts to things. And so he's a great guy for me to consult with on just, "What do you think, what are the shareholders really looking for, what do we need to do," et cetera, "How would the market react if we do ABC," and he's great. And then we, more recently, added Naomi. And obviously, she has a lot of experience, and she comes with a different angle from the business side and the strategy side. So I'm so fortunate to have this Board. You have no idea. In some cases, my VPs, they came from other companies. And they came to me and they said, "You're transparent with the Board. You're not trying to kind of spin things around when you present it to them." And I said, "Yes, they're just great. They're working with me. We're open. They give me a lot of advice," and it's really awesome.

That's great. A couple of questions just about the discrete memory, Coby, and also the selector work -- or actually discrete memory without a selector. Can you give a bit of an update what's happening on that discrete front?

So it's ongoing. It's in the background. Right now, the big focus by far, and the team knows it very well, is let's get these agreements done, let's get some revenue, some cash coming from customers. And so that's really the big focus. Now we are continuing the work in the background on the selector and on looking at discrete chips, et cetera. It's a different business model. It's something that requires a lot of R&D work. And whenever I have a contention on the different machines in the fab at Leti, they tell me, "Well, what do you want us to run now, the selector side of things or finish work for one of the customers," and stuff? And obviously, the trade-off is clear. So this is moving on. And by the way, when I say selector, we actually have several options that we're looking at in different directions. I think I want to be cautious here, but maybe there's one other company in the world who's trying to find a better selector for ReRAM. I'm not aware of really R&D going on right now anywhere on a more advanced selector for ReRAM. I think in that sense, Weebit is definitely a pioneer. And we will get there. But right now, I really want to close these agreements. I really want to get the embedded side ramping up, and that's the big focus, and that's the priority. So unfortunately, you can't have both of them at top priority, and the selector is lower priority.

Yes. Makes sense. All right. We have 12 minutes to try to race through as many questions as we can. Obviously, agreements with customers and foundries is, I think, #1 topic. The second topic that comes up a lot is share price. How are you feeling about the share price, where it's at, and how it's performed?

Well, as I think we mentioned earlier, I expressed Dadi, Atiq and I collectively, I think, put in more than $100,000, we purchased shares in June. I think that reflects what I think about the share price. We have intensive shorting activity going on. I think it reflects the fact that the Australian market doesn't understand how semiconductors work. They think that the fact that it's taking this time to close deals is some sort of weakness or reflecting some sort of problems. And then even when we announced that we taped out DB HiTek, you see a rise in the share price and immediately, the shorters come in and they hammer it on the head and stuff. Weebit is a company, for me, that is a long-term project. Semiconductors, like I mentioned earlier, things don't happen overnight, but the potential of Weebit is huge, okay? Everyone needs nonvolatile memory. The world is based on semiconductors today. The world is based on smarter semiconductors, it's the smarter power management tools, the smarter automotive side, the smarter edge AI. Of course, we can start talking about AI and we can go on for hours. So there's really a big demand. We're seeing that the market today realizes that it needs ReRAM. I was talking about TSMC and Infineon and other companies, it's very clear now they're heading in this direction. The analysts are showing and reflecting this as well. We will be achieving these agreements. We're making progress. I haven't had any negotiation where the customer said, "Okay, let's just drop this," right? I mean once we engage in a real evaluation and negotiation, we're always progressing. So I'm confident we will be signing these deals. Again, I just showed it in the own way that I could with my personal actions. And the next year is going to be amazing. The next year is going to be really exciting with more agreements being signed. And at a certain point, I think the shorters in the market will realize that, hey, this is a big potential. And hopefully, the shorting activity will stop. Yes, that's what I believe.

Yes. A quick question here, what are the things the ASX makes you put in an announcement? Why can't you just announce everything you do? I can quickly just cover that off. Unfortunately, there are some materiality thresholds that dictate what can go on to the ASX platform. Anything that isn't material, which we think is with shareholders knowing, we will e-mail out to the distribution list. So if you're not on that distribution list, you can sign on to that through the Weebit website.

But I do want to make just a comment that today, our relationship with ASX is already much better. It's been improved. And I think it actually ended up being even visible. There was a period where there were some significant misunderstandings. I think we managed to overcome them. So it's a good opportunity to thank the guys at the ASX also that we managed to resolve issues. And I think we are aware now of what they want us and don't want us to announce and how and things are going to be better now.

Great. A question, what is the difference between Weebit's ReRAM and TSMC's?

Honestly, we don't really know. We know what they published in papers. We know what we hear in the market. But there's not a lot of public information. I think that overall, we are not very different. I believe we have some advantages over their ReRAM. It may be on the temperature side or other sides. But it's hard for me to say things definitely publicly because we're trying to learn more about their year-end. We don't know enough. But the mere fact that they are selling ReRAM is a huge benefit for us. This is really helping us a lot. They are creating the awareness in the market that, hey, ReRAM is it. Now that they published their road map, and they show the ReRAM is it, that's what they're focused on. Again, that's helping us a lot because it's lowering the resistance level of customers.

Great. Michael asks, how is Weebit going with scaling down the size of its memory?

We're constantly working on it. It's part of the work with different foundries. We demonstrated already we have the wafers at 22 nanometer. And as I mentioned, that demo is really a fascinating one. By the way, I think, next week, there's going to be the Flash Memory Summit or they changed the name.

The Future of Memory and Storage.

Future Memory and whatever. Anyway, as we progress in the testing and in the work on those wafers, we will be giving a little bit more data at that conference. And we are engaged with multiple companies at what's called the FinFET level in the teens. We have several PDKs of FinFET in our R&D. We've run simulations. We've demonstrated that we know how to manufacture ReRAM on FinFET. FinFET is very different from FDSOI, which is very different from standard CMOS. And we've shown that we know how to achieve actually very good results at the FinFET level. So it's a matter now of pushing forward with the foundries.

Got it. All right. Quick response here. Cash balance of $62.8 million as at the end of the 30 June, how do you feel about that and the runway that it provides you?

Well, I think if you do the math there, you can see that we have cash right now for more than 2 years forward. So we are feeling very good about where we stand. I'm not feeling any pressure right now in terms of cash, which I think is the most important thing. The focus is let's work on closing these agreements. Once we close the agreements, once we have more customers onboard, we can look at how do we move forward and grow the team and what is needed. But right now, there's no cash pressure.

Great. Another question, with Israel in war, what is the backup plan for Weebit? Where would the team work from? Where will the management work from?

Obviously, the whole situation here is very unfortunate, disturbing, frustrating. I have a lot of words and ways to describe it. But at the same time, and as I said right after October 7, Weebit has contingency plans. This is not, unfortunately, it's not the first, not the second and not the fifth war that I've been through in my life. The Israeli industry is, again, unfortunately used to this situation and prepared for it. We demonstrated it, I think, over the past year. We have our plans. And we worked with major consulting firms on BCP plans, on business continuity. So we're prepared for it, and we've demonstrated it already now. And by the way, since I joined Weebit, this is the third round of escalated activities, and we've been through them without any issues.

Okay. Great. A question here, Yole suggested in their latest 2024 report that ReRAM will gain 34% of the total embedded market, question from these attendees. What do you think about that, Coby? Are you comfortable with that forecast?

I believe ReRAM will gain much more than that. I think that Yole is looking at things. I mean there is a transition now in the market. And it used to be MRAM and it's moving to ReRAM. I think they're still looking. They're giving a very heavy weight to PCM, for example, where in reality, there's only one last company that's developing PCM right now. STMicro is the last one that's developing PCM. But STMicro sits very close to Yole. And I think there's an influence there, so they give a higher weight than they should to PCM. I think they're not taking into account that big transition that we're seeing now with TSMC moving to ReRAM. So I think it's going to be bigger than that.

Great. All right, last question. A few questions have come through around what's next over the next 6 to 12 months for Weebit. Or where do you see Weebit in 6 to 12 months? And my take on that is what excites you most on the next 6 to 12 months at Weebit?

We're going to have a really good time. We've been working so hard on pushing forward all of these evaluations. And we're finally getting to the point where the companies are seeing the results. We're making progress there. We're moving now with some of them to the negotiation phase. I mean, for me, the next 6 to 12 months, we need to see some agreements signed. We need to see the initial revenue ramp and moving forward, getting even product companies, really working hard to see some product companies engaging. I mean, that's really the key part. So I'm really full of energy with just working on all of that stuff and waiting for that to happen.

Awesome. All right. Well, we've now exhausted our hour. So thank you to everyone that made the time to join. We had a really good level of participation there. So thank you. Thank you, Coby, as always for making yourself available for shareholders. If there are anything that you don't think we adequately covered off or follow-up questions, please feel free to e-mail me. My e-mail is on the ASX releases or go through the Weebit website. But that's it from us. I'll hand over to you, Coby, just to close things out.

Well, thanks, Eric. Thanks for organizing this. And I want to thank especially, of course, all the shareholders that have been there and being patient with us. I know it's not easy. I know you're not used to the semiconductor market and the time scales of semiconductor. But I just can tell you we're making very good progress, and we're in a much better position today than we were a year ago. And I'm sure that in 1 year, we're going to be in an even much better position. So thanks again for your support and for your patience, and we'll make it happen.

Thank you. Good afternoon, all.

Thank you.