Mike Bishop: Okay. And a question, maybe you can comment on the replacement for TSI. I think you addressed it a little bit in the prepared comments, but has a replacement for TSI been signed?
Scott Bibaud: Yes. I would say we have spoken to almost a dozen companies, maybe less than that, but a lot of companies. And it seems very clear that we have a lot of good options. We have customers who – I mean possible suppliers that are much better process technology that they used to have at TSI, and we have ones with more specialty processes that we want. So, it’s – I think we are getting very close to starting working with one or more of them, and I don’t have any doubt that we will end up working with multiple suppliers, not just one, but I think we will get started with the first one very soon. So, yes, I think the replacement for TSI is well on its way.
Mike Bishop: Okay. Great. And then Richard Shannon had a follow-up question. Richard, if you would un-mute and turn on your camera.
Richard Shannon: Mike, it will not let me turn on my camera, but can you hear me?
Mike Bishop: Yes, we can hear you. Go ahead.
Richard Shannon: Okay. It’s not allowing me to, oh, there we go. Now, we can do this maybe. Yes. There we go. Scott, I wanted to follow-up on one of your responses to my earlier questions here related to STMicro. You said you basically showed them all the data that they had requested back in 2020, but they didn’t start until 2023 when they pulled it off the shelf. So, I guess kind of applying this to your other set of customers’ engagements, how many other customers have you essentially satisfied all of the data that they have asked for and are sitting around? And do you think it’s reasonable to think about a delay between having the technology like satisfying all the specs, and then waiting x number of years. In STMicro’s case, 3 years before you get to production.
Is that something that you expect, or is that an extraordinarily long time? How would you relate this to your experiences with other customers that we haven’t gotten to that point, but seem like you have made some good progress?
Scott Bibaud: Yes. It’s a tricky one to answer, Richard. To be honest, ST surprised me. I think surprised all of us. We gave them results. They were very happy with them. And then just – they just never went off the dime. We kept in conversation with them all the time, like every few months, we would meet with them, and they would say, yes, keep waiting. I don’t think that we – so we have a number of other customers that we have run wafers with, and they have seen good results, and they are not currently in the process of going to production. But I couldn’t say exactly how many or whether I think that’s going to be typical that they pull it off the shelf. I can say there are a lot of customers who we – yes, that we have shown the good results to, and we keep talking to them and they keep kind of pushing it off a little, but we do think that we will engage with them sometime soon and JDA1 is a good example, but we have other examples like that.
Actually, we have a lot of examples like that. So, yes, the frustrating thing is if they could just see the results and pull the trigger, that would be great. And that’s what – normally in a business like ours, a customer puts out an RFQ, they need a certain type of product, and you go and pitch your product and when they decide yours is the best you win it. That’s what my experience in semiconductors has been my whole career. But this business we have is a little bit different. We are going out and proactively telling them, hey, if you used our technology, we can make your product a little bit better. And for the most part, we are kind of putting that into their minds. So, even after we have convinced them, and that’s the case, we sometimes have to wait until the planned change of that process or node is happening for them to implement it.
And – but I can’t say whether that would take a long time or a short time, we haven’t had enough experience in it.
Richard Shannon: It seems like in all the conversations we have had on these conference calls and offline as well, it just smells like RF SOI kind of is mirroring STMicro in a way where it seems like you have suggested or outright told us that you satisfied the requirements. And I think once or more than once you have even talked about some requirements to changing at some point. But it seems like RF SOI mirrors a lot of what you described with STMicro, is that a fair comparison?
Scott Bibaud: Yes. And let me just – so, let’s say we have a customer, and we don’t engage with them and we engage with them as soon as we can, right, as soon as we can get them interested enough to start testing out our technology. And that customer had a plan for some process to bring out a new version of it in 2 years. And so, they are working on that new process. And simultaneously, we are trying to show to them that our technology is good. And maybe even with only 1 year left, they say, wow, your technology is really good. But it’s too late to implement into our other process. We have already been working for it for more than 2 years. And so, you kind of missed the bus on that one. Now, we have to wait for that one to go to production and then be in production for a few years before they make a new version of it.
And so sometimes that’s just what we face. But like I say, if we have enough shots on net, we are going to hit the right timing with some of them and the other ones will come back around.
Richard Shannon: Okay. Fair enough perspective. One last question for me, Scott, just touching on the topic of leading edge, and I think I have even asked this in past calls here, but it sounds like you have got some long period of engagement with multiple players at leading edge. I guess my essential question here is, do you think that work is mature enough here that you have the possibility of intersecting with the first generation of a new technology coming out. You are talking about nanosheet or gate-all-around here, which I think it’s being implemented first on the 2-nanometer node with one or more guys out there. Do you think you are going to be early enough to do that, or does that seem like it might be more of a follow-on derivative process later in the timeframe?
Scott Bibaud: Hard to say, I can say that the leading-edge guys know about our technology, and they know how it could help them. And so, it’s kind of – obviously, if one of them had decided they were definitely going to do that, they certainly would have had to do a license with us. One of the things about gate-all-around and the new nodes, they are so hard to make. They are so complicated that they can’t just – what we call a demo in the industry is when they run some wafers, and they send them to us and then we put our technology on and then we send them back. But for gate-all-around, it’s so complicated to make these things that we can’t do that. We really have to install at the customer site. And so when we do that, they have to sign a license.
When they do a license, we definitely will announce it. And obviously, we haven’t announced it yet. But those – I think we are well positioned to get into one of those. I also think the gate-all-around and even the most advanced FinFET nodes still are running at relatively low yields compared to like more mature nodes. And there is room for us to be even incorporated to improve yield in those designs. So, fingers crossed that we will get that done.
Richard Shannon: Okay. Fair enough. That’s all my questions again, Scott. Thank you.
Scott Bibaud: Sure. Thanks Richard.
Mike Bishop: Okay. And just one last question here from the Q&A line and that is, could you describe how the business model for entering the GaN market would be different from your approach to license – approach to-date to licensing of MST?
Scott Bibaud: Yes. So, the GaN market is very interesting. Actually, this whole compound semiconductor, it’s an area that we have been doing research on for a few years. So, although this is our first announcement, it isn’t something we just started thinking about. So, there is multiple ways that our business model can be used for GaN. We can just license our MST technology, just like we do today with regular semiconductor makers to adopt MST on wafers and then build GaN wafers on top of that. That’s one way that we could do it. We will probably license our GaN technology separately from the rest of our licensed technology because it may have very high value. The second thing we could do is maybe we could become a manufacturer of GaN wafers at least in modest volumes, and that would allow us to generate revenue.
And if it was a very high-value technology, maybe we could make a very nice-looking gross margin there and help to subsidize the rest of our business. We don’t have a decision to do that yet. I mean we have a very strong philosophy about making a business model that’s got a lot of leverage to the bottom line. And so, if we were to become a manufacturer, obviously, we would have to have a lot more CapEx, which might not work well with that. But we could do something on the smaller volume manufacturing side. And – but I think for our customers who would go into high-volume manufacturing, it’s most likely that we have licensed directly to them. So, we have been looking at the GaN market for a while. We have had a lot of marketing studies going on about how to approach it, but we haven’t got a final determination about what we do.
I can tell you, I think it’s a – it would be a much faster time to market. And I am pretty excited about the prospects of MST there and the prospects in a few other areas in the compound semiconductor market.
Mike Bishop: Great. Okay. Scott, why don’t – that concludes the Q&A session, if you could proceed with any closing comments.
