#71 – An Interview with John Edmond - Luciferous LED Lucubrator

Chris Gammell: Welcome to the Amp Hour.

Dave Jones: I'm Dave Jones from the EEV blog.

Chris Gammell: And I'm Chris Gammell from Chris Gammell's Analog Life.

Chris Gammell: And I'm John Edmond from Cree.

Chris Gammell: Whoa. Well, welcome, John.

Dave Jones: Hey, John. Thank you. Thanks for coming on our humble show here.

Chris Gammell: Thank you for inviting me. And so where are you located, John? Where is Cree out of?

Chris Gammell: Our headquarters is in Durham, North Carolina. Ah, so slightly warmer than Cleveland.

Dave Jones: Is that a North Carolina accent, I hear? No. No?

Chris Gammell: It's not.

Dave Jones: Where do you hail from?

Chris Gammell: Is yours a North Carolina accent? It's Australian. I hail from Western New York. Ah. Western New York? Really? Yeah, Western part of the state. Yeah, near Rochester. Oh, okay. I'm from Buffalo.

Dave Jones: Is there a different accent between East and West New York? Oh, there's a big difference. Oh, yeah. Oh, really? Okay.

Chris Gammell: Yeah. Yeah. Yeah. So me and John are both upstaters, right?

Chris Gammell: That's right. We're upstate New Yorkers.

Chris Gammell: Yeah. Yeah. I'm from a part with a really nasally A, too. I'm from Amherst.

Chris Gammell: I know right where that is.

Chris Gammell: Amherst, New York. Yeah.

Chris Gammell: Then you should have said, are you from near Rochester? Because I've had people say that to me. Really? Yeah.

Dave Jones: Wow. They can pick out that little nuance in the-

Chris Gammell: They can pick it out, yeah. It's been a while. Yeah. Wow.

Dave Jones: That's amazing. So what's your position at Cree?

Chris Gammell: I'm the Director of Advanced Optoelectronics Technology. Big, long title. Dang.

Dave Jones: Wow. And co-founder, is that correct?

Chris Gammell: I am a co-founder, yep.

Dave Jones: Co-founder. We have a co-founder of Cree on the show.

Chris Gammell: We have a co-founder of a huge company that makes a lot of cool stuff on our show. This is pretty great. I got to say, I'm very excited.

Chris Gammell: I'm excited, yeah. I'm very excited. I agree with Chris. We have some really cool stuff here. Yeah.

Chris Gammell: That's why I keep doing it. I love it. That's great. That's great. So you're in charge on like a technical level or, I mean, managerial? What is it these days? I think Ginny said that, Ginny, your social media person who's helping us out, said you're still at the bench a lot of times. Tell us about that.

Chris Gammell: I am still at the bench. I was soldering leads on a part today. So, yes, I have a group of guys that work under me. And our focus is to work on the next generation LED chip or component or light bulb or whatever. Wow.

Dave Jones: Would most of that be like silicon research or manufacturing process research?

Chris Gammell: I would assume it would be. Well, you said silicon. So we don't work on silicon. We work on…

Dave Jones: I stand humbly corrected. My knowledge of silicon… Well, you know, semiconductor manufacturing is nigh on zip, whereas Chris has actually worked at a manufacturing… A manufacturing…

Chris Gammell: Hey, Dave, it would have been a lot worse if you said silicone. Believe me. Silicone. Yeah. I've heard that several times. Well, we would have just signed up the show right then.

Dave Jones: Any of that manufacturing technology here in, you know, the backwater that is Australia, you know. Ah. So… Okay. Well, we've got one fab here which specializes in silicon on sapphire. Okay. Yeah. But that's the only lab that we have left here, and we've only ever had one or maybe two. So, yeah. Unlike the U.S., where you guys… Where are there a dime a dozen? Yeah. A dime a dozen, please.

Chris Gammell: But I'll answer your question because I think I know what you're asking. So, we're working on the chip side. So, that might be things that are in the chip fab, for example. New chip designs, different looking chips. They're all… All our chips are based on silicon carbide, however. So, we either leave the silicon carbide on or remove the silicon carbide and put it on another type of substrate. And that's… I'm talking about the epi layer. Right. So, the lighting-mitting layer is all done on silicon carbide, which makes us different than pretty much every LED manufacturer in the world.

Dave Jones: And why is that? Why are you using silicon carbide and nobody else is? Do you guys have it locked up in patents or what?

Chris Gammell: Well, that's part of it. But the other part of it is we vertically integrated. We started a company based on silicon carbide LEDs and silicon carbide electronics. So, that was our core difference, right? That was our ace in a hole. And so, no one else was able to do silicon carbide.

Dave Jones: So, that's what gives you the efficiency advantage in terms of efficacy, lumens per watt and that sort of stuff? We believe that.

Chris Gammell: Right. We still have the brightest parts, the highest lumens per watt, and compared to anybody on Sapphire right now. So, we continue to push the envelope and they're continuing to push the envelope and we keep staying ahead 5% to 10% along the way. Excellent.

Chris Gammell: So, a silicon carbide wafer, right? So, you said that's the… You use the epi layer at the top? Yes. I was under the assumption that the entire boule was… Or the entire wafer was silicon carbide. Is that incorrect? Correct.

Chris Gammell: The wafer is silicon carbide, but we grow a gallium nitride layer on top. Ah. Okay. The active light emitting region is indium gallium nitride quantum wells. Holy moly.

Dave Jones: Zoom straight out of my head.

Chris Gammell: Yeah, Dave, we're in our element here, buddy.

Dave Jones: Yeah, I know. Sorry. Back to resistors.

Chris Gammell: Back to resistors. I need an op amp.

Chris Gammell: So, but, you know, everybody grows in-gang quantum wells on Sapphire, but we do it on silicon carbide. Okay. That's how we started the company and that's what we continue to do.

Chris Gammell: And so, can you tell… So, sapphire then is different than silicon carbide, correct?

Chris Gammell: That is correct. It's a different material. Sapphire is aluminum oxide. Ah, okay. And aluminum oxide has a different coefficient of thermal expansion. It's got different properties that… Lattice mismatch is another thing. It's not quite as closely matched to the nitrides as silicon carbide is. So, we felt right off the bat that silicon carbide was going to be a better material.

Chris Gammell: Yeah. Wow. Yeah, I had heard that a lot of the silicon wafers you guys… Or silicon carbide wafers you guys make are actually one of the only sources in the world. Is that right? Of the actual raw wafer?

Chris Gammell: The source of… Yeah, we're the source… I would say we're the best source of the wafer. There's some… There's more companies coming out with silicon carbide now for high-power electronics. Okay. So, but we're still…

Dave Jones: So, do you manufacture the wafers in your own fab? Do you do everything, like bring in raw materials and… Yep. Yeah, you do everything. Have you always done that from day one?

Chris Gammell: Yes. We are vertically integrated from basically raw material, powders, gases… Right. …to finished light fixtures that you can buy in Home Depot. So, that's pretty vertically integrated.

Dave Jones: Sand in one end and cool stuff out the other end. You got it. That's awesome. Da-da-da-da-da-da-da-da-da.

Chris Gammell: Take a bucket of sand from the beach and… Yeah. You're pretty accurate because it's really sand and coal. Really, it's charcoal and sand.

Dave Jones: Yep.

Chris Gammell: Man.

Dave Jones: I know. It's brilliant. All done in the one factory. It's like… Yeah. It's almost cartoonish, you know. I'm picturing like a cartoon. Yeah, yeah, yeah. You know, they card it in one end and magic comes out the other end.

Chris Gammell: That's the idea.

Dave Jones: Oh, that's great.

Chris Gammell: Now, I didn't realize that gallium nitride was a critical part of the process too. I always thought that the LEDs you guys were making were mostly silicon carbide as a base and then you did like the usual dopants and growths on top of that. I didn't know it was necessarily a gallium nitride growth. Is that a bad assumption on my part?

Chris Gammell: When we started Cree, our first LEDs were all silicon carbide. You're 100% right. Okay. So, from 1988 to 1995, so for seven years, we made silicon carbide on silicon carbide-based LEDs. So, we were the only ones that had blue LEDs out. Then we started working on nitrides and we have a – nitrides are much more efficient, right? They can be, but there was problems in making those materials. So, for the first seven years, it was all silicon carbide and since then, it's been nitrides on silicon carbide.

Dave Jones: So, there's only been one major technology leap there?

Chris Gammell: That's a huge technology leap.

Dave Jones: Right. Oh, yeah, but – Yeah, just a whole new material, whatever. So, do you see anything in the – any major leaps like that in the future in terms of, you know, the actual process technology or is it more, yeah, I think we've got it. We'll just tweak these because we're getting near the – tweak the current processes because we're getting very near to the – what we can achieve or something like that.

Chris Gammell: I'd say the latter is true. I mean, there's not too many – there was a lot of work on zinc selenide, which is another direct bandgap material. And it just floundered because of degradation. This material is very rock solid. And we're approaching efficiencies now of 75% to 80%. So, you're not going to get a lot more out and – Right. I just don't think there's any other materials that are going to get close to that.

Dave Jones: Wow. Interesting. So, basically, just your job now is sort of extracting, you know, an extra – oh, let's – our goal this year is to extract a couple of extra percent out of this baby and keep leading the market. Yeah.

Chris Gammell: Is that what sort of stuff you're working on? Yeah. My slogan with my team is no photon left behind.

Dave Jones: Oh, brilliant. Are there signs around the company with that?

Chris Gammell: No. T-shirts? I should. T-shirts? We should get t-shirts.

Dave Jones: Yeah, absolutely. I think – Jenny, we need to work on that. Yeah, Jenny should do that. Jenny's shaking her head.

Chris Gammell: No, that's great.

Dave Jones: Send us a couple of t-shirts too, please.

Chris Gammell: Yeah, I'd take one. I would take one. Yes, please. We'll take a t-shirt. So, can you go back to those early days? Because I imagine – you know, we were talking about this factory, you dump sand in, and then you eventually get light bulbs out. But how do you start? Well, Lightbeast is vertically integrated. They're actually making modules, right?

Dave Jones: This isn't the Edison era.

Chris Gammell: This is – LED bulbs? Exactly. We just – that's true. Yes. Okay.

Chris Gammell: Okay. Well, regardless, when you're dumping the sand in, you're actually making the silicon carbide the first time. I mean, I've seen videos of them actually drawing at least silicon bulls out. We were talking with Jerry Ellsworth on here a while back about that process and stuff. But how do you – I mean, you don't just like go down to your store and buy, you know, a foundry that can do that. Like, how do you – do you just start out with really tiny ones and then tiny wafers?

Chris Gammell: How do you – Well, we started out with one centimeter wafers. Oh, man. So, that's pretty small. Yeah. So, you can imagine – that's really small. But you can't – the silicon carbide doesn't melt, so you can't pull it from a liquid like you can with silicon or sapphire or some of those other materials. So, it's very different to grow. Okay. And so, it's very different.

Dave Jones: But that's all you need to start out with, right, is like a one centimeter wide wafer because, like, that's all you need for experimentation. Oh, that's how – Right. Is there any advantage to go into larger apart from the yield in actual volume manufacturing?

Chris Gammell: Oh, sure. Is there any – Well, you say yield, but it really comes down to cost.

Dave Jones: A cost, yes. That's right.

Chris Gammell: Yeah. Yeah. You know, you could – I did my PhD on one centimeter wafers that we grew in a lab. So, no, you can do a lot of R&D on that. But to get to a real manufacturing process, you need, you know, four inch, six inch. Oh, of course.

Dave Jones: You'd need, yep, large ones. So, is that how the company started? It was based on your PhD research and other people's and you got together?

Chris Gammell: How did it – please tell us a story. Yeah, so we had a group of students at NC State University here in Raleigh, North Carolina. And we were all working on silt and carbide. It was an Office of Naval Research Fellowship – Office of Naval Research Program. And we basically were able to create enough technology that a couple brothers that were in the program that I graduate, Eric and Neil Hunter, they approached myself and another guy, John Palmer, another guy, Kelvin Carter, and one other fellow, Tommy Coleman, and said, hey, let's start a company. And that's really how it started.

Dave Jones: Right. How does it work? Because that was government funded. I assume being naval research funding, that would be government dollars.

Chris Gammell: It was government dollars.

Dave Jones: How do you – sorry?

Chris Gammell: It was definitely government dollars.

Dave Jones: Right. So, how do you then take that technology and then what are the legal aspects of taking that technology and just forming your own company? Because I assume all that technology would be public domain because it was government funded. That's right.

Chris Gammell: So, what you do is you basically – the first thing we did is we wrote 10 patents on the technology. So, all the technology that we invented, we wrote the patents. And then we asked – we basically asked for an exclusive license from NC State. So, we came out. We wrote the patents. We went to NC State and said, can now – after we just wrote these, can we have an exclusive license? We'll give you stock in a company. And they said, okay, we've got to go back to the government and ask them. And the government said, yeah, we don't care about it. It's just you guys aren't going to amount to anything. So, hey, it'll never work. So, it'll never work. So, we got exclusive license to the technology.

Dave Jones: Fantastic.

Chris Gammell: Wow. And they didn't think the stock was going to be worth anything.

Chris Gammell: Yeah.

Chris Gammell: But it was.

Chris Gammell: I mean, that's a zero.

Dave Jones: So, they still own stock? Penn State still owns stock in the – NC State. Oh, sorry.

Chris Gammell: NC State. No, Penn State's got some problems right now. So, let's not talk about them.

Chris Gammell: Yeah, different issues. Yeah. I'm right. Sorry. Different issues. Yeah. So, do they still have ownership in the company, though? Does NC State still?

Chris Gammell: I think that they sold all their stock. I don't know for sure, though. You'd have to ask them. I don't know.

Chris Gammell: Right. Okay. But they helped you get started in some capacity because they gave you rights to the patents that you worked on.

Chris Gammell: That's right.

Chris Gammell: It's interesting, though. And that kind of thing happens now, too. There's technology transfer offices at a lot of schools. I know Michigan has one. MIT has one. Stanford has one. Because they're great sources of new work. You know, that's how that transfer happens is a big deal. And a lot of schools are concentrating on it more now. I don't know if they had an office back then.

Chris Gammell: You're right. Now they are. Because back then, it was just like they didn't have that. They were like, I don't know what to do with this thing. MIT had it. But places like NC State, there just wasn't a lot of companies that came out of there. SAS was one of them. SAS software came out of NC State. So there was a precedence, but very few – they just didn't know what to do with it. So it worked out good.

Chris Gammell: Yeah. That definitely works out good. So tell me about the – I'm still stuck on this silicon carbide generation or how you first made it, right? I looked up silicon carbide for the show, and I researched a little bit before then even. And I was reading about moissanite, too. And so was ever the inclination to be like, you know what, I'm just going to go grow some diamonds in the lab? Has that ever crossed the mind? Growing diamonds? Well, I mean, moissanite is fake diamonds, right? I mean, or close to diamonds, right?

Chris Gammell: I would say it's better than diamond, but that's besides the point.

Chris Gammell: You have to say that, right?

Chris Gammell: Of course I have to. We're not working on diamonds, but silicon carbide, it turns out, has a property called birefringence, which means it has a different refractive index in different directions. So it basically has a better sparkle than diamond. But that was not why we started doing it. Well, of course not, yeah. That was not our dream. Right, right.

Dave Jones: Okay, cool. So how did the, what capital funding did you need and how did you get the capital funding to start up?

Chris Gammell: It was basically talking to friends and family, going to a bar and sit there and someone writes you a check for $5,000. And it was some other folks. It was all family and friends, basically.

Dave Jones: Right. So there was no venture capital?

Chris Gammell: Zero.

Dave Jones: Zero. That's amazing. That's fantastic.

Chris Gammell: Yeah. We came close a few times, but we didn't do it.

Chris Gammell: I'm sure the draw is there, right? It's like, oh, we could not eat ramen tonight, right? That's true. Yeah. Man.

Dave Jones: And I bet they're glad they did if they held on to their stock when it, did it float? When did Cree float? Was it a huge success? I don't know the history there.

Chris Gammell: Well, we went public. How many years after? We went public, I think it was 93 or 94, Jenny.

Dave Jones: Right. So that's like five years after you started?

Chris Gammell: Yeah. We started in 87. So it was like seven years, something like that.

Dave Jones: Okay. So that's a fairly long time to wait to get your return on your investment, your $5,000 you put in there, I guess?

Chris Gammell: Yeah. And if the guys held on, I mean, they became millionaires because the price was like very, very, very low. Yeah. Relative to what it went to. Yeah. Yeah.

Chris Gammell: I saw your interview with Jenny and about your early days and working day in and day out. Could you tell us more about just like how that kind of ramped up then? Because going from that $5,000, borrowing $5,000 up to having large-scale fabs is a big stretch. So I'm sure there's some good stories in the middle there too.

Chris Gammell: Well, like I said, like you said, we worked day and night. I mean, it was like come in and see each other at breakfast time and then go to lunch together, then go to dinner together, then go home at 10 o'clock and see each other again for breakfast. And not six days a week, seven days a week. Wow. So it was – we had to build – you've got to remember, we built every piece of equipment we had. We didn't – there was nothing out there you talked about. I know. We built everything. We built our epireactors because there wasn't anything out there. We built our crystal growers. We had to invent everything.

Chris Gammell: That is awesome. That is fantastic.

Dave Jones: Wow.

Chris Gammell: There was nothing out there commercial. Wow.

Dave Jones: So where did you get the boot – did you continue to bootstrap? Like did you sell early parts and get some revenue? Where did the money continue to come from over those years before you actually really sold commercial volumes and stuff like that?

Chris Gammell: As far as – before we went public? As far as – before we went public? Yes. Yeah. When did revenue start flowing in? It was like $11 million. Okay. Revenue started flowing in probably 1990. We did a – I think that was the first deal we did with Mitomo in Japan as a distributor for us and are still our distributor today in Japan. They – I think we were like two weeks from running out of money and we signed that deal. Oh, wow. It was I think a $2 million deal and they got exclusive distributorship to our LEDs in Japan.

Dave Jones: How many LEDs were you churning out in those early days?

Chris Gammell: Oh, I mean probably a couple thousand a month. It was like nothing.

Chris Gammell: A month, right.

Chris Gammell: It was nothing.

Chris Gammell: Wow. But they were a lot higher price back. I remember – I saw also in that interview with Ginny about the blue LEDs and how that was such a different thing when that came out that – I mean I'm sure that they got some pretty decent dough for them.

Chris Gammell: They did. I mean Osram was there. Siemens was a company that had them out there for – they were $75 a piece and you could barely see them load up. $75 for a blue LED. In 1987. Yeah. That's crazy. So we came out with – And that's $19.87. Yeah. So we came out with ours. What's that worth today? That would be probably $150, $200 or something. Yeah. Double that at least. And it barely lit up. It was terrible. So we came out with ours. We were under a dollar. We came out with $0.99. Oh, wow.

Dave Jones: Wow. Instantly. So you – like you instantly slashed it two orders of magnitude almost.

Chris Gammell: That's right.

Dave Jones: Wow. So that was huge.

Chris Gammell: And that's why I think Sumitomo bought on because they said, hey, Cree can do it for a lot less than everybody else. And we were about four or five times brighter than Siemens was at the time.

Dave Jones: Was there a temptation to sell it for $10 instead of $1 just because you could? Because that's what the market – the market was used to paying $70. Well, hey, $10 would be a revolution.

Chris Gammell: But they didn't sell. They sold them only for like medical applications for blood gas analysis or some crazy thing. We were going into indicators. We were going into actual indicators for electronics at that time.

Chris Gammell: Yep. So you're almost playing against the red and the yellows, right?

Chris Gammell: Yeah. Yeah. But for high-end stuff.

Chris Gammell: Okay.

Chris Gammell: And then our first big –

Chris Gammell: I still like blue.

Chris Gammell: Yeah, I love it. But the first big application was the dashboard of Volkswagen automobiles. That was the big one. That's where we really turned the corner.

Dave Jones: Oh, was that when – So you're responsible for the horrible trend in blue dashboards.

Chris Gammell: Yes.

Chris Gammell: So was that when they redid the Beetle, like with the trendy Beetle and –

Chris Gammell: Yes.

Chris Gammell: Was that the era of that?

Chris Gammell: That's it. Yeah. It was mid –

Chris Gammell: It would have been mid-90s.

Chris Gammell: Yeah. Yeah, I remember that. And then Jetta was real big too. And Jetta had that nice blue dashboard. I remember that too.

Dave Jones: What was wrong with just using a regular light bulb with a blue filter? Well, you tell me. What was the – How do you – Well, yeah, the bulbs blow, but they're still using bulbs these days. It's crazy. I can't believe it.

Chris Gammell: I guess they still do some, but I don't think much. Yeah. Most everything is LED dashboard.

Dave Jones: Yeah.

Chris Gammell: But the blue was – I think the CEO of Osram and Volkswagen got together. The two CEOs got together, and they looked at it, and they said, this is just beautiful. Nothing.

Dave Jones: Right. It was just a more beautiful, in quote marks, color than a crusty light bulb with a blue filter on the front. Exactly. Right. Exactly. Right.

Chris Gammell: Cellophane. And they said, we want to do this. And so that was our first big one. That's awesome. Yeah, cellophane doesn't work. Yeah.

Chris Gammell: Yeah, like those stage lights where they do the cellophane covers. Oh, yeah. Like 1,000 watts.

Chris Gammell: Yeah.

Chris Gammell: Oh, boy. So what about – I mean, so those first sales, I mean, that's a big deal. Then what about the ramp-up then? I mean, you started – was it actually after going public that it was a bigger ramp-up? Because I'm guessing you probably can't sell a couple thousand a month to get in the cars.

Chris Gammell: No, no, no. But even before then, we were selling silt and carbide LEDs, not nitride-based, but silt and carbide and silt and carbide. We were selling millions of them. Oh, okay. A quarter. So, I mean, they were making display boards, and they were doing indicators, and there's a lot of applications. But then once the Volkswagen hit, that just – like I said, that just went to – took us to the next level, and we were doing several million a month above that. So it just – that was the big ramp right after that. Oh, wow.

Dave Jones: Where do you manufacture all of your LEDs these days?

Chris Gammell: Every wafer is made here. Every epi layer is grown here in Durham. USA. All our wafer fab is here in Durham.

Chris Gammell: USA.

Dave Jones: USA, yes. And is that a deliberate choice by the company? Have you looked at the financials and gone, oh, it might be cheaper to offshore it?

Chris Gammell: That's a good question.

Dave Jones: But we'll keep it here and be patriotic.

Chris Gammell: There's a big part of our operation that's in China because they're – frankly, they're our biggest customers. So you have to be our biggest customers. So we bought Kotko, which is Chinese optical technical company, whatever it stood for. They were the biggest LED manufacturer in China.

Dave Jones: So you bought the biggest LED manufacturer. Ah, who were they? You take them over. Yeah.

Chris Gammell: But that was three or four years ago. Three or four years ago. So it's – or maybe five years ago. I don't know what it's been. But by being there, we have – we don't do any of our – any of that part of any of our alcohol critical technology. All that's kept here.

Dave Jones: Which is the wafer type stuff.

Chris Gammell: Which is the wafer part. That's right. So they take the wafer and they cut it up and they encapsulate it, package it, whatever you want to call it. And they go into LEDs for display boards like New York City. They go into everything after that. So they do a lot of great work over there. And we've got 2,000 people working over there.

Speaker ?: Wow.

Chris Gammell: Fantastic.

Dave Jones: How many people total in the entire Cree organization, the monolithic organization?

Chris Gammell: I think we just – about 6,000. Oh, wow.

Chris Gammell: Fantastic. What about some of the packaging technology? So, I mean, you talked about packaging done in China. But it seems like that's a pretty critical part to most LEDs because drawing that heat away from the dye and trying to – Getting the heat out, yeah.

Chris Gammell: Right. But – okay. So doing that is extremely important. But it's all – all the pilot lines are here. Okay. So we develop all the packaging technology here in Durham. And then we send – well, actually, I shouldn't say all of it. Some is done there. We have a good R&D group there as well. So we do some of it there. Most of it's done here. And then we transfer it over. Okay. Okay.

Chris Gammell: So what is in the future for that kind of thing? I mean, is it actually going to get into fancier package types for the actual LED packaging? Is it going to be external to that with better heat sinks on – you know, in a bulb configuration? Is it going to be – or just lower power draw on the actual dye, the silicon carbon dye?

Chris Gammell: I think it's going to be all those. I mean, right now, the standard packages are small. They're 3.5 by 3.5 millimeters or 5 by 5 millimeters or 10 by 10 millimeters. Will they get to be, you know, huge? I don't think they're going to get much more. I think some of the biggest arrays now are on maybe 30 or 40 millimeter copper core boards, basic on-board technology. Okay. So, yeah. But then, like you said, everything goes – it's got to get – you've got to have a heat sink that gets it to the outside world and then in the light bulb application. So all three things are going to evolve.

Dave Jones: Are there any plans for Cree to branch out into manufacturing full bulb replacements as in the whole package? Or do you guys just want to stick with the LED itself?

Chris Gammell: So right now, we actually sell bulb replacements. Oh, you do? Okay. Oh, yeah. Oh, I haven't looked. Okay. There you go. Par 38s, par 30s, six-inch can replacements. Oh, you do? Oh, yeah. We sell – like I said, we go from raw material to go into Home Depot. Home Depot, Home Depot. I don't know if you – do you have Home Depot in Australia?

Dave Jones: No, we don't. I heard it's big in the U.S., though, yeah.

Chris Gammell: It's just a big, you know, home department store, basically. Hardware-type store. Yeah. Hardware. Yeah.

Dave Jones: Yeah.

Chris Gammell: And they've got lots of lighting in there. There's a lot of LED lighting, and we're in there on their shelves.

Dave Jones: All right. I'm going to ask the big question, when do you see the critical mass happening where LEDs will take over from CFLs?

Chris Gammell: Completely take over? Or just the tipping point? The tipping point.

Dave Jones: The tipping point where it's – you know, it just becomes cost-effective and the longevity is better and yada, yada, yada.

Chris Gammell: Right, right. So today, you can get, I think, the cheapest bulbs out there, and they're not very good and they're not creased, but you can pay about $0.03 per lumen. So that's not really good because if you have to have 1,000 lumens, let's see, they're 430 lumens for about $10. Right. So that's the best I've seen. I think that's about $0.03. So if you could get down by an order of magnitude, which I think is possible, believe it or not, you could get to a dollar bulb for $430, then I think you're there. Because CFLs are giving them away. I mean, they're literally giving them away. So it's hard to compete with someone giving bulbs away, and that's what the power companies are doing.

Chris Gammell: Oh, okay. Okay. So you're saying it's almost subsidized at that point. It is subsidized. Yeah. Okay.

Dave Jones: So how many years away? How many years away? You're the best guy in the industry to guess this. Everyone wants to know.

Chris Gammell: I would say that you'll be able to buy a sub $5 bulb in the next two years. That's about a 500 lumen bulb.

Chris Gammell: Fantastic. Dave's setting you up for, he's trying to say, he wants to be able to point at something in two years and be like, aha, John said it wrong. He won't say that out loud, but that's what he's going for.

Chris Gammell: I think I'll be right. It might be lower than that, but I think I'll be right.

Dave Jones: Yeah. I think you'll probably be right too. I have no idea.

Chris Gammell: But for a dollar bulb, it's probably going to be a while, probably five years. Huh.

Dave Jones: So where does, hey, I'm jumping in here, Chris. Okay. I got in first, I think. All right. Where do you, what is your best Cree, what's the best efficiency on or efficacy on your Cree LEDs now compared to CFLs and fluoro tubes in general?

Chris Gammell: Okay. So. Is it on fire better? Are you talking about R&D or are you talking about.

Dave Jones: No, I'm talking about ones you can buy. I'm talking about commercial.

Chris Gammell: What you can buy right now, you can buy, I think our highest bins is about 170 lumens per watt. Yep. So that's about 2X the best CFLs you're going to get.

Dave Jones: Right. What about the larger fluoro tubes as opposed to CFLs? Like the strip.

Chris Gammell: They're like, the best ones are like, they're 100 lumens per watt.

Dave Jones: Okay. Right. So you guys are almost doing 1.8 times that.

Chris Gammell: And that light is a tube. So light goes up. They don't count. You're going to have losses, right? Because you have reflectors.

Dave Jones: You've got to have reflectors. Yep.

Chris Gammell: Right. LEDs are more directional. So you don't have, you can have a lot less loss in a system. Hmm.

Dave Jones: I've got a question that where, when did, I noticed the market years ago went from spec in LEDs in candelas and it moved to lumens, lumens per watt. Can you tell us the history of why that happened or how that happened?

Chris Gammell: Well, I think because we went from blue and green and red, right, to white. As soon as you go to white, people want to know how many lumens it is. Because that's what you have on your, right, that's what's on your light bulb.

Dave Jones: Yep.

Chris Gammell: Box. And how do you measure that?

Dave Jones: You measure that with a sphere, don't you? It's measured with a big sphere.

Chris Gammell: Yes, integrating sphere. That's correct.

Dave Jones: And integrating sphere. How do they work?

Chris Gammell: It's basically a, it's a big ball. You put your light bulb inside of it. And the inside is basically 100% reflectivity. Right. So it's just a big, it's a white painting and white coating. And then that gets detected in a, in a, a radiometric system and it measures the amount of lumens. It's, it's straightforward. They've been doing it forever.

Dave Jones: However, how many, how much fudgery, I don't know if that's a word. Is that a word? How much fudging is there in the industry with these figures? I mean, because a lot of them are bullshit. I mean, you buy these cheap ones on eBay and they're just absolute bullshit.

Chris Gammell: Yeah, you're right. They're going to, I guess they have a lighting fax now on most of the new bulbs to be an energy star and lighting fax. Lighting fax.

Dave Jones: Oh, okay. Is this a U.S. government thing? Yes.

Speaker ?: Is it?

Dave Jones: Yes.

Chris Gammell: Right. And it requires the lumen number, the color rendering index, which is the light quality, the power input. So basically you can get lumens and lumens per watt and color, warm white. It has a color temperature on there, warm white, cool white.

Dave Jones: Yeah.

Chris Gammell: And so you got to.

Dave Jones: Is there still a way to cheat that system though? Fudge your numbers a little bit here and there?

Chris Gammell: I don't think so. I think they're pretty accurate now. Right. If it has a lighting fax on it and if you look for, if it says energy star, then it definitely, and our bulbs do, then you definitely, there's no fudging there.

Chris Gammell: Right. Got it. And so there's a third party that does that kind of like evaluation then? Is that the idea?

Chris Gammell: That's correct.

Dave Jones: Okay. So it's not an in-house number. Well, we have an in-house system. Oh, of course you test them in-house so you know what you're going to get.

Chris Gammell: Right. Everything is sent out and qualified. Certified independently. That's correct.

Chris Gammell: So excellent. If I can go back to the LED bulb takeover that Dave was asking about. Yep. I mean, are you guys scaled up to actually take that on? I mean, are we going to always be buying Cree bulbs or is there actually going to, is there any way you guys could even do that in terms of sheer numbers?

Chris Gammell: Well, the biggest, yeah, that's impossible. I mean, you've got the biggest lighting manufacturers just throwing all their money at LEDs. That's Osram. Yeah. That's Phillips. That's GE. Yeah.

Chris Gammell: One's up the street from me, actually. Which one? Oh, GE. GE. Yeah, GE in East Cleveland.

Chris Gammell: So, I mean, all those companies are also, you know, they're good customers of Cree's. So, no, you're going to have a bunch of different bulb companies and the big ones are going to be the biggest. Now, we hope to be right there as far as the best quality. We're always going to innovate and try the best light bulb out there.

Chris Gammell: Yeah.

Dave Jones: Who's your biggest competitor at the moment?

Chris Gammell: Are the raw days, right?

Dave Jones: In terms of the raw, you know, lumens per watt.

Chris Gammell: Raw, okay, so on a raw chip side, which is really blue, just a blue pump, I would still say Nachia is the closest to us as anybody.

Dave Jones: Right.

Chris Gammell: Okay. And then after that, it's basically Toyota Gasai out of Japan, and then it just moves down to Taiwan Inc., basically. Huh.

Dave Jones: But you think you're back in the best horse with the silicon carbide, huh?

Chris Gammell: Yeah, absolutely.

Chris Gammell: Right. There you go. And so, the blue LED, though, that, I'd like to ask about the makeup of white LED. I mean, that is mostly a blue LED, is that correct?

Chris Gammell: That's, most applications use a blue LED pump, and they put a phosphor on top of it. And that phosphor glows either yellow or yellow and red, if you want warm white. So, you have the blue plus the yellow or green and red, and you get the full color spectrum.

Chris Gammell: Man.

Dave Jones: That's why it was easy for you guys to transition from the blue LED technology into the white stuff, which you're mostly known for these days.

Chris Gammell: That's correct. Yeah. But the other thing we did was our lighting systems, almost all our lighting systems have what's called true weight technology, which has a yellow LED, basically a blue with a yellow pump, and then we add a red LED to it. So, it's got two different types of LEDs. It's got a blue and a red and a phosphor. And that gives it very, very rich color. The color rendering or color quality, it looks like that of an incandescent.

Chris Gammell: See, that's kind of tough for me only as a consumer. You know, as an electronics person who has put white LEDs into designs and everything, I got to say, I'm like, ah, it's just a light bulb. I know there's a lot of nuance to it, but that's what I think of, right? I mean, I'm still stuck in incandescent days. I mean, that's a horrible thing, but it's just a reality.

Chris Gammell: Well, the key is the reason that our bulbs have been so successful is when they install them, people don't know it's LED. That's the key. Yeah. If you walk in someplace and you can't tell it's an LED, then if it looks like an incandescent or standard lighting, you've just won it. You've just won. And your payback is less than two years, and your energy costs went down by 80%. That's where you win. See, now that's awesome. That's where you win. Yeah. Yeah. Yeah.

Dave Jones: That's great, too. That's saving the world right there, right? See, I'm building a new lab at the moment, and I want to fit it out with LED lights, but it's still not even order of magnitude close to being cost effective compared to fluorescent strip bulbs. I mean, I did the math, basic math, and the payback period at the moment was 15 years or something. It was crazy.

Chris Gammell: So how do you think we're selling? How would you answer to the question, why are the lighting business growing fast as anything here at Cree? So if it's not cost effective, why is Los Angeles putting in 150,000 streetlights right now with Cree LEDs in them?

Dave Jones: Oh, no, because they- LA has high taxes? I don't know. But I'm talking about stuff that I can buy for my office. It doesn't appear to be. There are tons of applications out there where LEDs, every street, every traffic light in Australia has been changed to LED. That was five years ago or something. Right. There was a huge, big difference. But yeah, I don't know. Perhaps I've done my math wrong, haven't done enough research, but yeah, just for the sheer fitting out my lab at the moment, the payback was only two years, I'd certainly do it.

Chris Gammell: Are the lights on 24-7?

Dave Jones: No. No, they'd only be on seven, eight hours a day or something. Okay.

Chris Gammell: Yeah, data doesn't work that much, though.

Dave Jones: No.

Chris Gammell: So your payback, it's not 20 years. Jenny, what do you think that would be? Just off the top of your head, six years, five years, something like that? Yeah.

Dave Jones: I don't know where you're getting... Well, it was basically the cost of actually buying these strip replacement LED that fitted in the existing sockets. You know, that's where you're paying your high price, I think. Yeah, yeah, yeah. So, you know, if I totally ripped out all of the ceiling lights and then installed, you know, the cheapest fitting I could get with the cheapest replacement, then I could probably do it.

Chris Gammell: And if you're doing new construction, you need to go on our website and we have a troffer series that will fit into all those applications, but it's not replacing the bulb. It's putting the whole either two by two or two by four troffer up in that space.

Dave Jones: Yeah, yeah. That's the thing. I was pricing that based on the current available price for the just, you know, you flip in that, you know, you take out the old bulb, you put in the new strip fluoro bulb, the, you know, the four foot long one. Yeah. Kind of thing. And yeah. That doesn't work. They don't seem to be there yet.

Chris Gammell: No, no, I agree with you.

Dave Jones: And there's lots of nasty ones on the market that are just horrible. They're using these five millimeter, you know, one hung low brand Chinese, you know, the cheapest five millimeter LEDs they can get in there. And the, and the, and the design of them is terrible and they don't meet safety standards and they burn out and they're, ah.

Chris Gammell: Oh yeah.

Dave Jones: So what's the name of that company again?

Chris Gammell: One hung low, you said? One hung low. Okay. I'll look that up. Make sure you don't get those. Yeah.

Chris Gammell: Yeah. I mean, John, do you see it being, I mean, did you see Cree over the longterm being that vertically integrated though? I mean, where you guys are actually selling fixtures and light bulbs and everything, because I would think over the longterm, it's not going to be realistic that, you know, eventually you're going to have to just sell LEDs because that's what I would think because it's.

Chris Gammell: No, no, it's the opposite. We, someday we won't be selling LEDs. We'll be selling. Really? Well, again, if you go to our website, we sell LED troughers, we sell LED bulbs, we sell, that's the fastest growing part of our business is the LED lighting. We just, we just bought a company up in Racine, Wisconsin called Rude Lighting and it's all street lighting. And we have 800 people up there that all made 100% in the USA and we're, we're doing LA, we're doing street lights all over the country. So I would hope someday we would just be making everything for internal use. That would be, I mean, maybe it's five years from now. I don't know. 10 years from now. I don't know.

Chris Gammell: So you don't see the fab as a, as a longterm model. I mean, is that kind of the, I mean, not from the business side, but just from a, you as being in the business, you think it's the, the fun stuff's on the, the application side more than the development of the, the bits side, not the bits.

Chris Gammell: No, there will still be development. There's no question about it. We're going to continue to make better. And it's, it's the bottom line is you've got to decrease the cost of the, of the chip. So your, your dollars per lumen has got to go way, way, way down into the dirt, like 0.002 maybe. So two tenths of a cent, one tenths of a cent per lumen. If you can get to that point, which I think we can in the next couple of years. And then on top of that, you increase the lumens per watt as high as you can. So you can imagine if you could have three bucks in LEDs and a thousand watt bulb, you could make a bulb for five or six bucks. That's a thousand lumens. And it's not, it's not that far away. It's not that far away.

Chris Gammell: So just to give a perspective on that too, that 0.001 cent per lumen, is that, is that roughly what an incandescent is? I mean, I, I kind of take incandescent is like the bottom of the barrel.

Chris Gammell: No, I think, I think incandescents are a lot lower than that. Okay. Wait a minute. Let's say, so, so let's say you get a four pack of bulbs, a hundred, a hundred watt bulb, which is going to be pretty soon, but they claim they're 1500 lumens, right? So let's say you pay a buck for that four pack. That's 0.25, that's 25 cents divided by 1500. That is 0.000, that's not, $2. So it's two hundredths of a cent per lumen.

Chris Gammell: Wow. Okay. It's not, they're nothing.

Chris Gammell: They should just give them, they're nothing. But, but they're, they're power hogs. I mean, they, you know. Oh yeah. Very. Oh, of course. Very power consuming.

Chris Gammell: Yeah. Yeah. Okay. Well, what about some of the other stuff that, I mean, Cree makes, I, we had some questions from listeners too. Alan asked, you know, he, Alan, one of our, one of our ham friends was asking about non-LED stuff. So like the power devices for GAN and MOSFETs I've seen before. I mean, what's the, what's the plan for that kind of stuff? Because I think that's a lot of our listeners are interested in that stuff too.

Chris Gammell: Well, we just released a MOSFET recently. Yeah. Um, I think it was a 600 volt, um, are, we've had a Schottky diode out and I don't work in that area anymore, but I, I know a little bit about it. Um, so the Schottky diodes, uh, are out there, they've been out there for a couple of years. So there are a lot of power supplies and a lot of solar, solar systems, solar, solar inverters. So that, that's coming along. I think the next set will be the higher voltage versions of everything. Um, so stay tuned. The power people need to stay tuned. Okay.

Dave Jones: I had no idea you guys were working on that sort of stuff.

Chris Gammell: Oh yeah. Definitely. There you go. It's a reasonable part of our revenue. Yeah.

Chris Gammell: I, I saw, I saw the N channel MOSFET and I know this isn't your area, but I, I just, I was really excited when I heard about it. I saw the N channel and I saw the, the announced higher voltage stuff and I got really excited about it. And, uh, I don't know. I, I hope there's like P channel coming and I, you know, just all the different types of FETs you can get. That'd be really cool. But I know that.

Chris Gammell: I'm not, yeah, I'm not sure. I think most of the stuff is N channel because mobility, the mobility. The mobility is so crappy in this with holes. Yeah. Yeah. In general.

Chris Gammell: Right. Right. Right. Yeah. It's always harder to get P channel stuff. Right.

Chris Gammell: But I haven't worked on a P channel MOSFET here in 20 years. So it's, I have no idea what they're doing. Come on, John.

Chris Gammell: Throw the, throw the bunny suit on. Just get back in there. It's true. I do. I wear the bunny suit all the time. Oh, I don't miss that.

Chris Gammell: I stick with the two terminal devices.

Chris Gammell: Oh, that's good. That's good. Uh, let's see some of these other, uh, uh, listener questions too. Um, oh, well, what about, what about like, uh, DC power for LEDs? I mean, obviously a lot of offline switching happens within the bulb for, for modules, but I mean, do you see a future where we're maybe with solar panels and using some of those Cree, uh, shot keys, you know, having a, having a solar panel, actually, I guess the shot keys are an inverter, so that doesn't really matter. But, um, but just, uh, all DC system, have you seen anything like that or any kind of request for that?

Chris Gammell: I mean, maybe, maybe Edison was right. And DC is going to be the final solution, you know, because you can imagine having solar panels on your house and writing, running DC everything. Cause AC was for distribution.

Dave Jones: It's the distribution. It's not for, it's not as good for local use. Yeah.

Chris Gammell: And that would, and that would really fit into the MOSFETs being there for low on resistance too. Cause then you have those drops. So exactly.

Chris Gammell: And of course, LEDs love DC. I mean, that's what they, that's what, that's how they work. So if you could make your, you can do your refrigerators, you can do everything else. You can imagine having independent, you know, power systems for your own house.

Dave Jones: I don't know about, we've probably gotten to the point where probably at the point where the, uh, the reliability of these new LED bulbs is dependent upon not the LED itself, but dependent upon the DC to DC converter circuitry and how well you design that.

Chris Gammell: You are a hundred percent correct. Most failures in those bulbs that you buy, um, are the capacitors. It's all, it's, it's, it's everything inside the ply. So they're getting better at it. But if you want a cheap bulb, some of those power supplies are just horrible. Yeah. Like those $10 bulbs. No. Yeah. No. You're going to, you're going to get probably, you know, one out of every 10 after a few, a few weeks will be burned out. That's the problem.

Chris Gammell: Right. And that's always heat, right? I mean, I know that's, that's even the problem for CFLs where you have these ballasts, these tiny ballasts inside, you're hanging out CFL upside down, it heats itself up and then there goes the cap.

Chris Gammell: They die.

Dave Jones: And that gives the whole LED industry a bad rep, doesn't it?

Chris Gammell: Sure. I mean, if you buy an LED bulb and you put it in and it goes out, you're saying, oh, the LEDs are no good.

Dave Jones: Right. This LED technology is garbage. I don't want that.

Chris Gammell: No, this electrolytic is garbage, right? Yeah.

Dave Jones: Exactly. Are there any government initiatives or anything to try and, you know, stop that sort of thing? I don't know if you can, you know, it's, you know, it's a free market. If you want to buy your $2 replacement LED bulb from China that's designed with the lowest common denominator parts, then you're free to do that.

Chris Gammell: Right. That's why Cree has a reputation it does because our bulbs are more expensive, but they don't die. They're dimmable. They have true white color. They look like an incandescent, blah, blah, blah, blah, blah, blah. But we have something called Energy Star. If you get an Energy Star emblem on your part and it's got the lifetime on there as well, and that's part of the Energy Star certification, then you've got a part that's going to work. So you look for, again, you look for the Energy Star logo. Chances are you're going to have a good bulb. Nice.

Dave Jones: Speaking of LED life expectancies, what is the trade-off between the operating current, you know, like the LED might be rated for operating at 350 milliamps. What's that, you know, what is the trade-off between lifetime and heat and temperature, maximum temperature and all that sort of stuff?

Chris Gammell: Can you give us some ballpark? Sure. I mean, it's... Rules of thumb there? The rule of thumb is you want to have a low thermal resistance to your part. So you want to have it down around 5 to 8 degrees C per watt for the component itself. Then after that, you want to rate it, you want to look at the maximum rating and probably not run it there, but you probably want to run it down near, you know, half of that or something like that in general. And then you have to be able to get the heat taken away. So the heat syncing is critical. And it's not the LED, it's not the chip. It's the packaging material around the chip that usually has trouble. All these are encapsulated in silicones. And the silicone... Silicone, huh? The silicones. The silicones, not silicon. Right, right. You wouldn't see anything. Silicones have a very high, you know, transparency for optical transparency. But they also have issues with lifetime. But they're much better than epoxy. So really, you're balancing all that. So the best thing you could do is have a high lumen per watt component that you're not creating a lot of heat. I mean, think about it. If you're 100% efficient, then you're creating zero heat.

Chris Gammell: Yeah. Yeah, that's right.

Chris Gammell: That's not going to happen. But if you're at 50%, then you've got a much better situation than if you're at 10% or 15%.

Dave Jones: What's your efficiency at the moment? What's your best efficiency for the best one on the market?

Chris Gammell: Okay, talk about white or blue?

Dave Jones: White.

Chris Gammell: The highest lumens per watt part, I think I had mentioned, was 170 to 175 lumens per watt. Now, if you look at efficiency, that's about complete, including conversion of blue to white, which is you have a loss called Stokes loss. Stokes loss is about 20%. But if you take all that, they're about 50% efficient.

Dave Jones: Right.

Chris Gammell: So 50%.

Dave Jones: How much do you see that improving in the next couple of years?

Chris Gammell: I think you'll see over 200 lumens per watt in the next couple of years.

Chris Gammell: So Stokes loss, is that like something you can't get away from basically based on the – is that just the light absorption of the phosphor?

Chris Gammell: Exactly. Okay. That's exactly – It's just basically your dominant wavelength of your chip over your – so the emission of your chip, let's say it's a 455. That's what normally the pump is. So 455 divided by your emission of your phosphor, which is roughly 555. So that's roughly 80%, 75%, something like that. So the other part's lost. It's gone.

Dave Jones: So what's the maximum theoretical efficiency that you'll be able to achieve?

Chris Gammell: I think you can get up – the maximum theoretical is – I think it's roughly 230, 250 lumens per watt or something like that.

Chris Gammell: Right. Okay.

Chris Gammell: Wow.

Chris Gammell: And there's a law that governs all this stuff. Isn't there like a Moore's law analog basically? There is. I forget what it's called for LEDs.

Chris Gammell: Well, but it's basically – you can calculate if you have 100% blue pump and you have Stokes loss, then what you're going to get. Now, if you mix red, green, and blue, and if they were all 100% efficient, then you could get lumens per watt over 300, like 350 lumens per watt. Oh, okay. Right? Because you can get white with red, green, and blue, but now you've got to have the right ratios of light, and it's got to be 100% efficient and all that jazz. But that's how billboards are done. I mean, full-color displays are all done with RGB.

Chris Gammell: Yes. I looked it up. It's called Heitz's Law. I was talking about over time. Heitz, that's right. Heitz's Law. Sorry, yeah.

Speaker ?: That's right.

Chris Gammell: And that's like how – basically how this is increasing over time. It's steady increasing over time, basically.

Chris Gammell: Yeah, they said it would double every – I can't remember what the number is.

Chris Gammell: Yeah, I can't find it here. Oh, well. It's getting better. Oh, every 18 to 24 months.

Chris Gammell: Yeah, but eventually it comes down to dollars, okay? If you get to 200 lumens per watt and you can do it for a penny, well, okay, well, that's great. If it costs you $10 to do it, it doesn't work. So it's lumens per watt per dollar. That's the key.

Dave Jones: That's the one that's going to win regardless of how many lumens per watt it actually is. Yeah. You've got to throw in the dollar.

Chris Gammell: If it's going to be real, yeah. If you want somebody to buy it.

Dave Jones: Yeah, yeah. That's right.

Chris Gammell: Well, speaking of dollars and China too, I mean, you mentioned China before as they're being held off by the ENERGY STAR standard. But we've talked on this program before about them controlling a lot of the – what do they call it? The rare earth metals and a lot of the rare elements. Is that affecting you guys in terms of being able to produce LEDs and the more rare chemistries? Not at all.

Chris Gammell: Not at all? Okay. No. There's still a lot of rare elements here in the U.S. too. And I wish I knew what all of them were. I think indium – you need indium, gay, nitrogen. They're all here. And then some of the other ones for the phosphors. There's some – there's definitely some in the U.S. But that's not – there's enough out there.

Chris Gammell: Okay. Okay. Yeah, because we usually saw it for like a lot of the electrolytics. Like tantalum is a hard one to find. And then the magnet side of things is neodymium.

Chris Gammell: Yeah, that stuff's very rare. That's what they call them rare earths, right? Yeah. Right.

Chris Gammell: Yeah, that's right. Okay, so a lot of the process – I always thought indium was in that group of rare elements. But I guess I was wrong about that.

Chris Gammell: Well, you know, early on they were saying, oh, there's going to be an indium. There's going to be a shortage of indium because they use it for a lot of electronics. There's plenty of indium out. Okay. Yeah.

Dave Jones: So how many LEDs – do you have any clue how many LEDs you're churning out per year, per month these days? Billions. It'd be astronomical.

Chris Gammell: I can tell you it's billions. I don't know the exact number. Billions. Billions. Wow. That's crazy. Billions. Yeah, it's a lot. It's a big – That's brilliant. No pun intended. It's a big number.

Chris Gammell: And how big of wafers are you guys working on these days? Is it 6-inch, 8-inch, 12-inch?

Chris Gammell: We have 4 and 6 in production, both 4 and 6. Okay.

Dave Jones: Is there any advantage to going to larger wafers for what you guys are doing? Eventually.

Chris Gammell: I would say, yeah. We usually make a transition to a larger wafer every, you know, 3 to 4 or 5 years, something like that. So our major production is on 4 and now a lot of it's going to 6. And the next stop will be 8.

Chris Gammell: So do you guys get a lot of the secondhand gear from like the Intels and the Samsung?

Chris Gammell: We definitely look for good prices out there.

Chris Gammell: Yeah, I was always surprised. I used to work in dry etch and I was surprised, you know, like you could just get new kits and I mean, if you don't mind old screen, you know, like an old control mechanism for it. It's really not that big a deal, you know. It's still doing the same thing.

Dave Jones: How are individual parts actually tested at the manufacturing stage? Are they batch tested, individually tested? And what's involved?

Chris Gammell: At the chip level. On a wafer. Yeah. We check it on a wafer.

Dave Jones: Right. Okay. So you actually probe it with an automated machine and how does it test it? Does it test it for just sheer light? Is it a go, no, go light output or do you actually characterize?

Chris Gammell: So it's got, it's the wavelength, it's voltage. So it's electrical and basically electrical and optical. And then once we slice the part up and put them in a package, we test those also.

Dave Jones: Yep. Right. How long does it take to individually characterize each one on the actual wafer? At the wafer level. How fast is it? Milliseconds. Milliseconds. Yeah. Fantastic. Fancy.

Chris Gammell: Yeah. So you can do a lot and you don't just have to do one at a time. Think of.

Chris Gammell: Okay. Yeah, of course.

Dave Jones: You can have a multi-probe machine that comes down with multiple sensors for each one and then multiple. Exactly. You know. Exactly. Drivers and, yep.

Chris Gammell: Wow. Okay. And so on a six-inch wafer, how many chips or how many die will you get out of a six-inch wafer on average?

Chris Gammell: Are you talking about one millimeter die or that's, or smaller die? Oh, yeah.

Chris Gammell: I guess that matters.

Chris Gammell: Oh, it makes a big difference.

Chris Gammell: Yeah. Yeah. I guess on your, how about on your high brightness processes? I mean, like those are probably the biggest die, right?

Chris Gammell: Yeah. So on a four-inch, I'm just, I'm just thinking out loud here, about 8,000 die, 8,001 millimeter die on a four-inch. So it's two and a quarter times that.

Chris Gammell: So, yeah. So it's going to be roughly 16, 20, 16, 15, 16, 17,000 die. Now that's. Wow. Okay. Wow. But for our smallest die, the smallest die, we're up to about a million die per wafer.

Dave Jones: Wow. Oh, wow.

Chris Gammell: Yeah. That was always, that was always kind of the, you know.

Dave Jones: Okay. You can't probe all of those at once. Huh? That's one hell of a bed of nails.

Chris Gammell: It takes a long time to probe a six-inch wafer with a. Wow. We sell a, it's called a UT190 and it's 190 by 190 microns and it takes like days to probe. Pretty interesting.

Dave Jones: It takes days?

Chris Gammell: Yes. To get to every LED?

Chris Gammell: Yeah, it's a million.

Chris Gammell: Huh. That's a lot of die. That's a lot.

Dave Jones: Okay. So you actually manufacture one wafer with a million on there and then it takes a day or two to test it? Yeah. Wow. How many, how many test machines have you got? I mean, I assume you'd keep churning out the wafers because you don't want to sit there, you know, you'd be stockpiling those wafers. How many test machines would you have going in parallel to?

Chris Gammell: Oh, we've got hundreds of. Zillions. Hundreds of testers. Hundreds of testers. But anyway, I mean, that was always kind of the, the futuristic thing to say, hey, someday we'll have a million LEDs on a wafer. Well, we got, we got there.

Chris Gammell: We're there. We're there. Next up, a billion. Yeah.

Dave Jones: How important would it be to improve your testing processes to lower in the final cost? I assume that'd be a big component of it. How much would testing cost compared to a raw manufacturer kind of thing? Would you know that?

Chris Gammell: The good thing is testing equipment that we've had here is paid for. So, you know, the cost is basically zero and you put it on there and it's, it's all, it's all automated equipment. Yeah. So you just let it run.

Chris Gammell: Yeah. Yeah. So that's. And then just up time at that point. It's up time.

Chris Gammell: Right. Yeah. So that's, that's a pretty small part of our cost. I can tell you that. Fab life. Okay.

Chris Gammell: Right.

Chris Gammell: Yeah. Wow. Oh, fantastic. But anyway, there's a lot of, there's a lot of cool stuff.

Chris Gammell: It sounds like, it sounds like a really exciting time. I mean, I'm sure it's been exciting for the entire time you've been there, but it, I mean, just, just with the uptake of energy efficiency and, and the lighting, you know, the cost coming down to a point where it's, it's starting to approach that, that tipping point. I mean, it does sound really exciting. It is.

Chris Gammell: I mean, you know, I've been doing this since 1987. We started Cree and I'm probably more excited now than I ever have been. But I say that every, every, every time I talk to somebody, cause I am, that's, that's why I'm still here. I love it. Yeah.

Dave Jones: Are all of the original founders still there?

Chris Gammell: Uh, we have, uh, four of the original founders still here.

Dave Jones: That's pretty high percentage. That's good. Definitely. Definitely. Yeah. That's fantastic.

Chris Gammell: Not only that, but, um, three of us live basically within a thousand or 2000 feet of each other. So we not only, we not, we not only, you know, enjoy each other's company at work. We enjoy each company, each other's company all the time. We're like brothers, basically. That's awesome. Fantastic. Yeah. When you grow up together like that, you become very close.

Chris Gammell: Yeah. Yeah. You find out a lot about your coworkers when you're there at the lab at nine o'clock together every night on a Sunday. You got it. Yeah.

Chris Gammell: You got it.

Dave Jones: So you're, you're still a hands-on guy. You're still soldering, soldering, not soldering. Yep. Soldering stuff every day.

Chris Gammell: Uh, not every day, but I was today.

Dave Jones: Fantastic.

Chris Gammell: I couldn't, I couldn't get a damn mission to do it for me fast enough. So I said, I do.

Dave Jones: And you don't trust any bastard, right? You got to best job is to do it yourself, right?

Chris Gammell: That's part of it too. Yeah. Yes.

Dave Jones: Have you, have you got hobbies outside of Cree or does Cree consume your life?

Chris Gammell: No, no, no. I, I, I love to, uh, we're, we're doing a lot of traveling now. Um, and I, I play in a band. I've played in a band. I've played music all my life. So why is there dog barking?

Chris Gammell: Chris will get excited about that. Ah, sorry. There's a, there's a dog at my house. There, uh, is there a wolf or is there a dog? No wolf. It's just a, it's just my, my dog. Yeah. Okay, cool. Such a professional radio show. Here we have you. She, she always tells us right when I ended the show. I think we're actually our, uh, we're going to get towards the end.

Dave Jones: Our amp hour is up. Yes, it is. Our capacity is over. But, uh.

Chris Gammell: That's cool. Yes. Well, that's very nice.

Dave Jones: Thank you very much, John. That was fantastic.

Chris Gammell: It's so great to hear about LEDs and, and from a market leader like Cree, that's pretty, pretty exciting. So.

Dave Jones: Well, you're very welcome. It's a great insight into the company and how it was founded. Yeah. It's fantastic.

Chris Gammell: It's been a great ride.

Chris Gammell: Well, we look forward to hearing more from you on, uh, on the LED front. Definitely. And, uh. Okay. Great. So. Excellent. People can find Cree on Twitter at, uh, at Cree. And, uh, I think LED Lighting Revolution or something like that. What's the website, John?

Chris Gammell: Oh, LEDRevolution.com. Any of them works. Yeah. LED, LEDRevolution.com, CreeLED, or Cree.com, whatever.

Dave Jones: Okay, cool. And do you have a personal Twitter where you tweet you'd like people to follow you or you're not into that sort of jazz?

Chris Gammell: I don't even know what Twitter is. Oh, right. Okay. I've had a flip foam until recently.

Chris Gammell: Right. But it had a badass LED on it, right? Right. And then it kicked off the wheel of it. Somewhere it did. Yeah. Yeah. Yeah. All right, John. Thanks so much for being on the show. Really, really appreciate it. You're very welcome. Take care, guys. Thanks.

Chris Gammell: All right. Bye-bye. Bye-bye.

Speaker ?: Thank you.