Gayn Erickson
Analyst · Scott Asset Management
.: Those DiePaks and WaferPaks are proprietary and patented concept that we had implemented literally over the last decade. That allows us to make extremely fine pitch contact with all of the die on a wafer, or the small contacts on a singulated die or a module. These modules, these tiny devices and we carry more on trade shows, I mean, they’re really small, a couple of few millimeters or so on a side. So they’re extremely hard to handle and we do that with automated equipment that we also supply. So when you go to buy a test cell from us, you’re actually buying an entire solution that’s made up of the tester and burn-in system with all of the electronics. I said WaferPaks or DiePaks, which contacts the devices under test and a liner or an autoloader that does automated handling of all this, so you’re not doing any of this by hand. Those – the test system is configured for a typical application. But it’s still general purpose, in that it can be reconfigured depending on the device just through software with generic resources and all. The physical contact of the device through the contractors that we call WaferPaks and DiePaks is a physical routing to get to the X, Y, Z locations specific to the type of the device. So for every new device, you actually need a new contactor. So there’s two reasons people buy a new DiePaks or WaferPaks from us. They’re buying more or they have more capacity, so they buy additional capacity from us. And in that case, they would be buying both the test system and the DiePaks or WaferPaks. And the other scenario is, a year later or two years later, their devices change from flavor A to flavor B. And if the pin out or anything changes about them, then they actually throw away their WaferPaks and DiePaks that were used before unless that product is still in production, and then buy new WaferPaks and DiePaks from us. So the question is how many turns is that, or how often will that be? And at least, in a lot of the semiconductor space that I’m familiar with, there’s kind of a few categories. There was always the typical consumer devices that might have a, two to three-year lifespan or something along those lines, and then there was automotive, which lasted eight or ten years. And so, theoretically, on the consumer guys, every couple of years or so, they might replace those devices with new probe cards or in our case WaferPaks or burn-in boards, or in this case DiePaks for our test systems, okay? Now the piece of it that’s really changed is, I think, it was just maybe six months ago just going through this discussion with someone when they pointed out that, if you think an automotive application is eight years old, go sit in a car right now. And if you look at the dashboard, there’s nothing that looks eight years old about it. In fact, I’m not sure what’s moving about as fast as the dashboard and the electronics that are going in automotive devices. What I know about cars that are going out next year with respect to these ADAS systems, forward-looking radar, even the biometric sensors. I drove an automobile in Europe a while ago and they’re available in U.S. as well, but there’s – the Mercedes now has it, where it’s continuously monitoring your face. And if you blink for too long, it shakes the wheel to ensure that you’re not dozing off to sleep, and I actually did that, so I remember it specifically. So there’s all kinds of things that are going in the automotive that maybe that turns even faster. The other piece is, on the whole mobile side of things, gosh, I don’t know, my kid’s phones don’t last very long. And if they do, we were replacing them before they die anyhow. And I don’t know the last time somebody did a breakdown, or tear down and saw the same part in anybody’s phone. So does that mean those devices have one-year life cycles, or two-year life cycles. So, the good news is the shorter the life cycle the better for us. And as we have an installed base out there of 10, 20 whatever systems at some point not only are people buying new systems, but they’re also going to be replacing those consumables. And we’ve specifically put in place capability and capacity to supply a very large number of them. So we’ve told customers that we have the capability that with lead time, et cetera, and the infrastructure of our supply chain that if we need to build 100 WaferPaks a month, we can do it. Now if you contrast that to the largest probe card suppliers, it’s in the same order of magnitude as the high-end probe card suppliers. So we’re optimistic about that. So one thing – another thing to think about is in – this isn’t something new if you look at the package part burn-in market. For example, the estimates that are out there for just standard package part burn-in systems, while the testers themselves might only have a couple of hundred million dollar market like a total available market, the actual consumable on the burn-in boards is closer to $1 billion –$800 million to $1 billion depending on the year. So the consumable has grown to be four or five times, several times larger than the annual spend, because people keep those darn burn-in systems around for decades. And that’s our experience with our installed base. But today, we do not supply consumables on our package part burn-in systems. But not only do we supply them on our wafer level and Singulated Die and Module, but actually we’re the only ones who supply it. It’s a captive supply relationship at this point. So I hope that gave you plenty of answer there.