They talk about how important it is for kind of a greener planet because of the efficiencies associated with both silicon carbide and gallium nitride and how those markets are going to grow significantly and not only displace some of the IGBT, but create new markets that didn’t even exist. So, my feeling is silicon carbide — we’ve had a chance to sit in front of customers in Asia and Japan, Europe, US, there continues to be a ton of excitement around silicon carbide and not just in the EVs. And guys, we’re not pivoting. EVs are critically important and they’re a huge opportunity for Aehr Test. But there’s a lot of companies, in fact, the last two customers that placed orders with us, they were not aimed at the EVs. I said that already publicly.
They were actually at other applications. And so those companies are making their whole strategy related to — so there is — it feels like a little bit of — silicon carbide was so much around its initial focus, which was, of course, with Tesla model 3 and the invertor, which kind of created the whole segment. But it’s like once people have made the investments to put the silicon carbide out there, there are all these other applications that would benefit from it. And to some extent with more suppliers, and I don’t want to say pricing reducing as much as it is just availability of it. There’s a lot more people getting into it. And so, silicon carbide is the real deal, it’s going to certain applications, all the way through the end of the decade, even with all the investment that’s going on.
Now silicon carbide, kind of the over-under, we’ve been using this 1000-watt thing. At a 1000 watts and above power transfer, silicon carbide is going to dominate. But for applications of 1000 watts and below, which could be individual servers, microinverters, and solar, certainly all the little chargers for Apple, your devices et cetera, they’re all going to be silicon — they’re all going to be gallium nitride, that’s what we think. And that’ll completely shift away the idea that you’re going to have a little transformer that plugs into the wall and heats up and it’s the size of your fist to power your computer, that’s all going to be gone. There won’t be any of those things in a few years. Albeit, these gallium nitride with — like a solid-state transformers or solid-state converters done through high-speed switching of gallium nitride and to some extent also silicon [Technical Difficulty] I think on the low end, not less than 1,000 watts, GaN’s going to win and those things are important.
Now, your next question was discrete versus module. This is where we have a show-and-tell on it. We don’t do these things with video feeds, but discrete normally refers to a single device sitting in like a package form with the two or three leads from it. Okay? This discrete device, if it fails, it only costs you the packaging. And so, the world had been actually testing those devices in the package form before we got here. And when people said that they’re going to put up to, say, 32 of these devices, 10 of these devices, 24, 48, we’ve seen all kinds of different ones. They put them in multiple die in the same package, we refer that as a module although technically it’s still a package, but a module meaning, multiple die inside of it. The yield loss associated with that becomes a real problem.
And so, if you had 30% yield loss because you had 30 die in there, I mean, that’s pure cost. Your manufacturing cost would be increased by 30%, your output would be decreased by 30%. So that was really what started this wafer level, why it became such a big deal in silicon carbide was because companies were moving to the module level. We still see that transition, that’s, like I said, a very public announcements by BorgWarner that’s all the US-based guys and more, [FZ] (ph) out of the European ones, the Danfoss folks out of Europe. Volkswagen has publicly talked about this. There is multiple companies the folks out of — starting to come out of what’s going on in Japan and also out of Korea. We think that the bulk and what we’ve been told by our customers, the bulk of electric vehicles and the big power conversion will all be modules.
And so, when they go to modules, they have — it’s so much more cost-effective or even a must to do the wafer level burn-in of those before you put them into modules. And so, that has been the leading indicator that was driving our business. And so, now the question is, did something happen? Have things slowed down to those cars, et cetera? I mean, clearly, we’ve seen some push outs of roadmaps and stuff from the likes of GM and Ford. I don’t think anybody is confused. But I’ll tell you what, you don’t see that — is that offset by the pull-in of Toyota, who last February was still maintaining they didn’t believe in electric vehicles until the CEO was displaced. We’re seeing — if you get outside of the US, you can just sort of — you can see the roadmaps for people.
I mean, there are companies like — Hyundai just announced, they just shut down two engine plants, it’s like the old Viking days of burning the boats. They’re not going backwards, they’re getting rid of that. In the US, you can’t do that because of unions and things like that. So it’s a big contrast and we think that modules are coming, they’re going to be from the largest to even some of the smallest automobiles that are out there, and EVs.
Tom Diffely: So even though it’s on everyone’s roadmap, would you say today it is very, very early in that transition?
Gayn Erickson: I believe still, I do. I mean, not everybody is doing it and — I don’t want to get into anything other than what is publicly described out there. But even some of the shifts with respect to what’s going on in Tesla’s business has had an implication, because I think people now, and again, I’m not saying anything that’s not public. Tesla which started this segment, they actually used a form of a discrete device in their current inverters. It has two die in each device and you can call it a module. And depending on the supplier, they can either do it with package form or wafer level. And some of those shifts could have an impact on us in the near term, but although I don’t — I’m not providing any insight into what Tesla’s plans are long term, but the prevailing thought process is that, by the end of the decade and certainly before that you’ll see all silicon carbide engines and modules or most of them.
