Credo surprised the industry at OCP by pivoting from microLEDs to microVCSELs for its first active light cable. This raises a fundamental question for AI interconnects: should data travel on a few fast lanes or many slow ones? Austin and Vik break down the physics, materials, and tradeoffs of both microLEDs and microVCSELs.
This episode is presented by Crusoe.
Crusoe’s Serverless Fine-Tuning trains an open model like Llama, DeepSeek, or GLM on your own data, and you pay per million tokens, not per GPU hour.
About $2 buys roughly 5M training tokens on smaller models.
Get $5 in credit to start: https://semidoped.com/crusoe
Things we cover:
Credo’s OCP announcement and the micro-emitter pivot
The physics of microLEDs (spontaneous emission, chromatic dispersion)
How microVCSELs work (stimulated emission, gallium arsenide)
The “wide and slow” parallel interconnect strategy
Why microLEDs might still have a future for scale-in
The role of materials: GaN, GaAs, and InP
Mentioned in this episode:
PicoJool’s Al Yuen: The Case for GaAs VCSELs in Scale-Up Interconnects (Semi Doped episode)
A MicroLED Field Trip to Avicena (Vik's Newsletter, paid)
MicroLEDs vs. Lasers: The Linewidth Tradeoff (Vik's Newsletter, free)
Understanding Energy Bands in Semiconductors (Vik's Newsletter, free)
This podcast is lightly edited for clarity.
OCP and the Credo Surprise
Vik: Credo bought Hyperloom, a microLED interconnect startup. Then at OCP this week, Credo’s first active light cable turned out to use micro-VCSELs. Why buy one technology and pick the other for your first product? The bigger question is how we connect AI systems. Do we keep pushing each lane of light faster or spread the data across hundreds of slow ones? And if we spread it out, LEDs or VCSELs? Austin and I are going to explain both and who’s building them. Let’s get into it.
Austin: Hello listeners and welcome to another episode of Semi Doped. I’m Austin Lyons of Chipstrat. I write a newsletter for institutional investors and tech executives and I also consult with semiconductor companies.
Vik: Hey, I’m Vik Sekar and I write Vik’s Newsletter on Substack and I’m the founder of Semi Exponent, a boutique research firm on semis. How’s it going, Austin?
Austin: Hey man, it’s going well. Just got back from OCP. Where are you at?
Vik: I am still in San Jose in a friend’s garage. I actually fly out to Japan today, purely out of holiday reasons. This is not for semiconductors. I’m not hunting Japanese company alpha. I’m not doing any of that stuff. We have great people already working on that. I’m literally going to go check out Japan for a week with my family on the way back because it’s on my way back to India. And so we’re recording this before I hop on the flight. But yeah, this is post-OCP. It was an insane show. There were so many people that they had to close down the registration and people were buying entrance tickets on the black market. And the halls were overcrowded and the exhibition halls were full, very exciting show.
Austin: Yes, it was something. I had a ton of fun talking to people on the show floor, talking to people outside. A lot of people ran into both you and I since we were there together and said, “Hey, we listen to the pod.” So thank you people who listen. It was fun to see you. I honestly didn’t sit in that many talks. I sat in some, but there’s just so many people to talk to, so much to see and also the talks were being put on YouTube. So I thought, yeah, I can watch them later when I get home at 2x, you know.
Vik: Yeah, if you go to the Open Compute Project YouTube website, you’ll see all the talks there, totally free to watch. I think it’s amazing that they did that pretty much on the same day. It avoided a lot of FOMO for the people who were there. And people like us decided, let’s talk to more people, let’s visit the show floor because they’ve wheeled in so much equipment and if nobody sees that, it’s a tragedy. And so we spent a lot of time on the show floor talking to people and that was so much fun. Now I have all the YouTube videos as does everybody else and so we can go and watch it at our leisure, at our correct spacing, whatever, right?
Austin: Which I do have some time blocked next week to catch up on them because there were so many good talks. OCP did a good job. Couldn’t get through them all. But one big talk and the topic for today, we’re going to do a micro-explainer on micro-emitters: microLEDs, micro-VCSELs, or both. Lots of “micro” to talk through. That was really the talk of the show. Bill Brennan, CEO of Credo had a keynote very early on. I think it was on the very first day and it was surprising. So let’s just dive into maybe let’s dive into the Credo news, but then we’ll zoom out and talk about micro-emitters, microLEDs, micro-VCSELs and everything. But yeah, do you want to give us the news first and then we can zoom out?
The Case for Micro Emitters
Vik: The whole stir-up happened on the conference show floor where everybody’s like, “Oh, did you catch that talk by Bill Brennan, Credo?” And I was like, “Oh, wait, I was watching the stuff over here.” But luckily, the YouTube videos helped. I watched it immediately as soon as it was posted. So the whole thing was the background is that for the longest time we’ve been thinking that microLEDs are going to be the way that cables are made with light because copper is reaching its limits on what it can do in terms of reach at a given speed. Optics has its pros and cons. It’s expensive, it’s power hungry, all of that stuff. So there’s always this middle ground where it’s a battleground for what kind of technology has to go in. And for the longest time, a lot of people were working on microLEDs. And these LEDs are pretty much like your light bulb in the ceiling, but it’s a small condensed version of it. And there are many companies working on this. Avicena; Credo acquired a company called Hyperloom for $92 million with this technology. Microsoft was working with MediaTek, I believe, on what they call Project Mosaic. And Marvell was working with this company called Mojo Vision to develop microLEDs. So all of these companies are artificial reality, AR/VR companies who had these microLED technologies already in production and they were like, okay, let’s pivot to data centers because there’s so much money to be made. So there was all this talk.
But then what happened at the talk at OCP is that everybody thought this is going to be microLEDs, or at least I did. But he got on stage and said, “Yeah, we were doing microLEDs, but we decided we’ll just do micro-VCSELs for the sake of reach.” And this sparked in everybody’s mind the feeling that micro-VCSELs as a technology is the future and microLEDs are kind of no more. That was the sentiment on the show floor, but I think there’s a lot more than that. I don’t know how much of it we’ll be able to get into. It’s really a long discussion. I plan to write an article on that as well. But yeah, let’s at least cover the basics of what these things are, why are we talking about these micro-emitters, what are the differences between microLEDs and micro-VCSELs and how to think of this going forward?
Austin: Yes, sounds good. I agree. Also, everyone was basically like, “What, are microLEDs dead?” You know, that’s like when we saw people that came right from that, “I guess it’s micro-VCSELs now.” But maybe also focus on Credo really quick and then we’ll get into the specific technologies. Anyone who’s been following Credo, obviously there’s this whole conversation over the past year of, is it electrical, is it optics? And Credo was the first to come out with the AEC, the active electrical cable, which was as we continue to talk at higher and higher speeds, the reach of copper shrinks and Credo said, hey, if we essentially put some digital processing on both ends and throw a little bit of power in there, we can extend the reach of copper. But still there’s as you look forward on the road map, the question is, how long will AECs last? And there’s long range communication over optics using very powerful lasers, for example, but this medium reach, there’s a question in the portfolio gap of passive copper, AECs, maybe something else, ALCs, active light cables, and then your traditional longer reach optics. And in this sort of intermediate area is where people were expecting Credo to come out and have these ALCs driven by microLEDs, but then, yes, the talk was about, “Hey, we learned a lot from microLEDs and actually micro-VCSELs are a good fit here.” So that’s again the framing, but now let’s get into it. So maybe start us with microLEDs, and then of course, I want to go into micro-VCSELs because people have probably heard a little bit about microLEDs over the past, but not as much about micro-VCSELs, although we have had a conversation with Al Yun of Pico Jewel, which was helpful and people can listen to that as well. But yeah, let’s dive into microLEDs.
Vik: Sounds good. I want to set the basis for what these are called micro-emitters and essentially, like you mentioned, there’s a reach gap. But I just wanted to mention one other thing about why this reach gap is important now because obviously lasers, indium phosphide-based lasers can go a long way and I want to mention why this material matters because indium phosphide is in short supply and everybody is scrambling to make this happen because optics is becoming a big thing in AI. But yeah, indium phosphide optics can go thousands of kilometers if you have to, but we don’t need all of that for the data center and the point is that why do you use a technology that is A) in short supply, B) capable of a lot more when you don’t need it? Can we use the right technology for the right application? Copper beyond 7 meters at 224 gig is about already pushing it. PAM6 which is a new modulation scheme is being considered to extend the reach of copper. So the copper work is going on on one side, but it’s becoming very difficult. Optics is like an overkill technology with indium phosphide. So there is a reach between say 5 meters and 50 meters where there’s a lot of applications you could do other things. And this is important because now it’s becoming way more important to expand the number of GPUs in the scale-up domain and it is becoming too hard to put too many GPUs in a rack. So even at the keynote from Nvidia at OCP, Ian Buck showed the picture of like eight racks connected together in an NVL576. And of course, that picture showed that it was co-packaged optics hooking them all up together. So that’s Nvidia’s solution, but everybody need not have to use co-packaged optics or near-packaged optics.
Instead there is a way to connect these things with other forms of light, right? And so 50 meters is enough to connect a lot of this stuff. And the question is, what is that technology? That brings us to micro-emitters. This is what the industry has been working on. And I’d say that although VCSELs have been around a very long time, right? Like VCSELs have been used in data centers. That’s one thing to remember about it, and we’ll cover more about this. But microLEDs, and we should enter into talking about microLEDs now, is that it is a far newer technology for interconnects. It has not been deployed in data communication applications. LEDs have been in displays and all of that stuff. But for communications, it hasn’t been a mainstay. Whereas VCSELs have been a mainstay of communication systems for a long time before we went to single-mode optics. But VCSELs have been around, they have been deployed, they do work at scale. With limitations, but they work.
Sponsor: Crusoe
Austin: This episode of Semi Doped is presented by Crusoe. They build the power, the data centers and the cloud behind AI. Did you know Crusoe can help you tune an open source model on your own data for just a few bucks? Frontier models are very good, but they don’t know your company’s terminology, your part numbers, or how you like your reports formatted. So, you keep stuffing instructions and context into the prompt of an expensive frontier model to get the job done.
But for simple repeatable work like reports, an open source model trained on your data can actually match a much bigger model. Think of a new grad with three months of on the job training. You don’t always need the PhD. And the open model is cheaper and faster to run. Fine tuning used to mean renting GPUs, working on orchestration and Kubernetes, writing training code and babysitting the entire setup process.
But Crusoe’s serverless fine tuning removes all of that. Upload your data sets, pick an open model like GLM, Llama or Deepseek, and they handle the training. You pay per million token used, not per GPU hour. And on the smaller models like Quen 38B, $2 can take you a long way with millions of training tokens available. Once the training is done, you can then deploy them on Crusoe Cloud in just one click or download the weights and run them anywhere.
Your data only fine tunes your model. And it doesn’t have to be work data either. You can try it on how you write emails. Pick a task, compare the tuned model to your current prompt and see how it does. Crusoe’s giving all new accounts $5 in credit to get started today. That’s enough to start tinkering with fine tuning right now. Links in the description.
How microLEDs Work
Vik: So, microLED. So, let’s talk about this new incumbent technology, the new technology that wants to displace optics. So, essentially what it is is that it’s literally a diode. Okay, if anybody’s done like a basic electronics course, it’s just like a P-type semiconductor with an N-type semiconductor.
It is made out of gallium nitride and it emits light approximately in the 420 nanometer range, which is actually blue light. And you can see it in these interconnects that if you look at the place where the microLEDs are, you can see blue light being emitted. So, it is blue LEDs that are doing this. You have red LEDs, you have green LEDs, anyway.
So, a lot of the work around microLEDs has been at 420 nanometers. And the thing with microLED light itself is that it emits light with based on what is called spontaneous emission, which means the moment you apply a little bit of voltage to this diode, it turns on. It doesn’t have a threshold voltage, like the minimum voltage after which it turns on. It’s like a continuously varying thing. It’s like the dimmer switch in your wall, right? The moment you start raising the dimmer switch, you don’t have to push it 75% of the way up for it to even turn on. Okay? This is an important analogy to remember because we need to use the same framing when we get to talking about VCSELs. But LEDs just turn on as you increase the light, just like you’re used to at home.
So, what ends up happening is that this is a different form of light emission called spontaneous emission where it’s just sending out photons randomly and they are not very correlated to each other. This photon decides to leave at this moment in time, the next photon decides to leave whenever it wants in time. They’re not coordinated, they don’t march together, they don’t do anything like that. Just leave. And then what are the other thing is there’s just like an LED in your home, right? This light is very, very broad angle light. It’s not like a pencil beam of light that you can use as a laser pointer. That is a pencil beam light, but LEDs are like dispersive. They give a wide angle. Okay? And so the third feature or bug, I don’t know, of microLEDs is that they don’t have a very good line width. So, if you look at it on a spectrum around 420 nanometers, they’re a pretty broad spectrum. And so this is important because as this broad spectrum light goes along an optical fiber, it faces what is called chromatic dispersion. And I know this is a lot of terms, but we have to cover these things.
Austin: Yeah, let me jump in here. So, if it’s supposed to be centered around 420 nanometers, when you say there’s like a broad dispersion, you’re saying some come out at 416 nanometer wavelength and some come out at 422 nanometer wavelength. So, that’s what you’re talking about here. So, the light coming out has different wavelengths.
Vik: Yeah, that’s what line width is. So, ideally you want if you draw the wavelengths of a light source across wavelengths, you want it to be like a single line. You don’t want like half a light bulb shape dispersion of wavelengths. You can’t have everything from 400 to 440 or something like that, which is approximately the bandwidth of this light source. It’s about 10%, I would say. So, you have anything from 400 to 440. So, this is what results in chromatic dispersion. And let’s get into quickly what that means. Chromatic means the color, right? Dispersion means how they scatter or something like that. They dispersed. So, what happens is if you send 400 nanometers down an optical fiber, it takes, I don’t know, let’s say one second. Okay? Now, 440 nanometers should also take one second so that they all arrive together, right? All the wavelengths should arrive together. Only then will it look like the same thing has arrived at the other end. But it doesn’t. Different wavelengths travel different speeds. So, what happens is like all these different wavelengths arrive at different times. And when that happens, when you send a pulse, that pulse has basically spread out in time. Which means that you can’t send pulses fast enough because now they start to overlap. Right? So, when you can’t send pulses fast enough because they start to overlap at the other end, that means your data rate limited. You can’t send data fast enough. You can’t keep sending bits because they’re going to start overlapping. So, you can only send bits slow enough so that you can still account for overlap at the end of the cable.
Austin: Gotcha, gotcha. So, if you have a very precise line width, it’s a lot easier to make sense of the signal once it propagates down the waveguide or whatever. But if your wavelengths are spread out, then they kind of smear and maybe overlap and by the time it gets to the end of the waveguide, it’s harder to tell. Was this a one or a zero? And to your point, then you can’t send signals quite as close together, which ultimately means you can’t send information as fast enough. Right? So, I’m assuming that you’re leading us to say that the problem with LEDs is that they can’t be used fast enough. Maybe that’s the problem for data centers.
Vik: Uh, yes. So, it’s a combination of speed and reach. So, you see chromatic dispersion limits how much spreading happens. The further you send this signal, the more the spreading. So, if you want to limit spreading, one way to limit it is that you use it only for like, I don’t know, 5 meters, for example, right? And then the spreading is limited. So, you can send a reasonably high data rate. Now, if you’re going to keep it like for 100 meters, it’s going to spread so much that your data rate, you can’t send data fast enough. They’re going to overlap, right?
Austin: Is that spreading, let me ask you really quick, is the spreading a function of distance because the photons are just bumping into the waveguide or something? Like, why does it spread more the longer you go?
Vik: But think about a cross-country race that happens for 24-hour cross-country or something. You have the cars coming at it much more different time intervals, you know, the leader might come in first, the other, the next car might come in like 10 seconds later, one minute later, things like that. It’s the same thing. Or you could think of this as a person running on a track, like the 100 meter dash has extremely close timings, but the 400 meter will have different spacings of people, and the marathon will have even more spacings.
Austin: Yes. Yes. Got it. Thank you. Carry on.
Vik: The other thing is that microLEDs fundamentally, you can’t turn them on and off fast enough to send data fast enough. Okay, this is a fundamental problem of the device. And right now we can only send about, I don’t know, maybe three to five gigabit per second on a cable of 5 meter reach, right? The LEDs can only turn on and off so quickly that you can send 5 gigabit per second. So, it’s limited in both directions, right? The LEDs can’t go faster inherently, and they don’t have reach. So, both ways.
Austin: So, if they can only go three or five gigabits per second, but we’re talking about a single lane wanting to, first there was 50G, then 100G, then 200G, and then eventually we want to go 400G. How do you get 224 gigabits per second if an individual LED can only send at 3 gigabits per second?
Vik: So, you have to bundle a lot of fibers together. So, you send many, many parallel lanes. For example, if you bundle about 300 of them together, you get a 1 terabit per second lane. I’ve written a very detailed post on all of this stuff when I visited Avicena’s office. We looked at the cables, we looked at the data that’s being sent. So, check that out on my Substack. I also have a full free article on Substack where I discuss this line width tradeoff in microLEDs and compare them to how lasers look, for example. That’s completely free, so if anybody wants to check that out.
But essentially, if you have such a slow data rate, you have to now put multiple parallel highway lanes together, and you have to put 300 of them, 1,000 of them, if you want to do both directions or one direction or what speed you want, you just have to double up the lanes. And people don’t really like that approach because it gets complicated to couple maybe 300 microLEDs into the cable. I mean, it’s doable, it’s not like it’s not doable. You have to put a lens to focus, because remember, I told you the light was dispersive everywhere, like it just fans out everywhere. So, you got to put a lens to focus it and make sure it goes into the fiber. Then you got to have a lot of fibers. It’s not an impossible task. The question is, do we really need so many lanes? What if we can do it without having so many lanes? What if we could use a technology that actually had reach? What if we use a technology that could modulate data at a rate higher than three or five Gbps? That is what brings us to micro-VCSELs.
How microVCSELs Work
Austin: Okay, so what are micro-VCSELs?
Vik: So, VCSEL is a way that people say it, but it’s actually an acronym. It says V-C-S-E-L. Okay, so that’s Vertical Cavity Surface Emitting Laser. Okay, so the L is a laser. This is not an LED. Okay, so it’s not surface emitting LED, no, it’s surface emitting laser. So, it’s vertical cavity. So, instead of a lot of the optics that you know in silicon photonics and all that stuff that people talk about, it actually emits from the side of the chip. So, vertical cavity means that the cavity is vertical and the light is emitting vertically. And our podcast with Al, Al explains very nicely how all of this works, and he has a prop in that episode. So, people should check that out. It’s definitely if you’re listening to micro-VCSELs because you will learn a lot from an expert on the subject. So, that’s definitely an episode worth checking out.
But let’s talk about this technology. So, VCSELs are made out of gallium arsenide. Okay. So, it’s not indium phosphide, it’s not gallium nitride, but it’s gallium arsenide. And VCSELs are quite commonly used in a lot of stuff, like Face ID is one example, and those are basically like lidar, light radar with light kind of thing. So, they just shine light on your face and they kind of identify you. That’s how it works. So, there’s a big supply chain for VCSELs because we’ve put them in billions of phones already, which is another benefit, right? Like I told you, this has been used already in data communications also. The supply chain is enormous. And on top of that, it uses gallium arsenide, which has been a mainstay technology in cell phones. Like every amplifier, transmit amplifier called a power amplifier in your phone is built out of gallium arsenide. Okay, so this technology is widely available. The material supply chain is amazing. You can build them on 8-inch wafers. So it’s not like indium phosphide, you’re limited to 3-inch and now people are struggling to get to 6-inch. No, but this you can build in 8-inch already. So, gallium arsenide is widely available. So, VCSELs are widely deployed in both communications and in phones and in lidar. So it’s like, wait, so can we use this technology instead? Like, why don’t we use this technology, right?
Comparing the Two Approaches
Austin: Got it, got it. So, okay, on the one side, we started with you have traditional indium phosphide lasers and there’s a lot of struggle manufacturing those. And on top of it, they’re very powerful and if you’re just sending light, even though it’s high speed, you’re just sending it a short distance between GPU racks, let’s say, it feels like it could be overkill. And then on top of that, you’re going down the route of co-packaged optics or near-packaged optics. So, there’s concerns about maintenance as opposed to just a traditional pluggable cable.
On the other end, you said, oh, well, we could use all these microLEDs and but now we’re going to take a 200 gig signal, we’re going to have to have some sort of gearbox to split it down into being able to send hundreds of 3 gigabit per second lanes. So, there’s a little bit of circuitry needed there and a little bit of complication. But then on top of it, now we have hundreds of fibers that we have to align to these hundreds of LEDs and they can’t go quite as fast. And so, you can start to see some manufacturing tradeoffs there.
But then you’re saying micro-VCSELs land in between. It’s still the laser technology, and it probably has a better line width and it can communicate faster as we’ll get into. But at the same time, that also unlocks the ability to not need to have so like hundreds of fibers, presumably. So is that where micro-VCSELs land is somewhere in the middle?
Vik: Yes, so micro-VCSELs are in the middle. It’s between microLEDs and indium phosphide lasers. So, the reach can go, I think the publicized reach at least now is 50 meters, right? So it can go 50 meters and it can do data rates of 50 gigabit per second easily. So, it can do 10 times a microLED data rates. It can do easily. It can probably do 100 gigabit per second. It can probably do 200 gigabit per second. Like people are talking about doing that stuff, although it’s very difficult because it has inherent limitations on how it works. But it can do 50. And if you look at a lot of the demonstrations that people, there are other startups working on this too. It’s not just Credo or someone else, but yeah, there are lots of companies working on this, but they all run it at like 25 or 26 or something. You don’t even have to run it at its capability and that has its benefits. Okay, so maybe I don’t go into that immediately, but it helps to run it slow. Okay. And I’ll contrast why this is later.
But essentially, you can choose the wavelength of this light to be either 850 nanometer, which is the widely deployed wavelength in all of communications. Like 850 nanometer VCSELs are well known, well studied. Okay. Then recently there’s been a lot of talk about 1060 nanometer VCSELs, which is what you can see from Lumentum, Coherent, all of these companies are working on 1060. It has its benefits. We let’s not get into the details.
Austin: Yes. And I will say one thing that stuck out to me about what Bill Brennan said was they can still get the same reliability. Even though these are lasers and everyone has concerns about reliability, Credo can unlock the similar reliability as copper because they’re already making an array of these micro-emitters. And hey, if you’re going to use, let’s say, 50 of them, why not have 55 of them in there and have some spares? So if one emitter were to fail, you could still have some extra fibers and some extra emitters and kind of on the fly always make sure that you have enough working emitters to be able to send the information.
Vik: Yeah, this is a feature that is common to both microLEDs and micro-VCSELs. If you have parallel lanes, why don’t you put a few spare lanes and you could always switch over to the spare lanes. So this is not really a feature that is specific to micro-VCSELs. It’s just wide but slow and this is the whole concept of the micro-emitters. We are going wide, many lanes, but slow. We are not going 200 gigs per lane like an indium phosphide laser. We are going slow. We are going 25. We are going 50. We are going even five or three like microLEDs. So whatever it is, this is slow, but we are using many lanes. And when we are using many lanes, you can always have spares. So this is not a unique feature, but it’s nice. Yeah, it’s a nice feature of wide and slow overall.
Austin: Totally. But and then I guess the difference with micro-VCSELs is you could increase this reliability from 7 meters all the way to 50 meters.
Vik: Yeah. Yeah, exactly. And that’s a good thing, right? Like you get a whole new domain of interconnect that does not use indium phosphide but solves a lot of problems. So futuristically speaking, it seems like that’s a better product option to build. And the fundamental reason I have to bring back one analogy I use. You know, I use the dimmer switch analogy. The reason that the micro-VCSELs can actually do a lot more than microLEDs is that in this scenario, when you increase the voltage to a VCSEL, it doesn’t emit any light. I mean, it does very, very weakly, but it doesn’t really emit any light till you bring it up to at least halfway through. And then bam, you got an explosion of light, right? That is the called stimulated emission of lasers as opposed to spontaneous emission in microLEDs, right? So stimulated emission means that all the photons now are in phase, like holding hands and marching like in a military regiment, forward and the line width is narrow because all of these photons have a purpose. They’re just marching, right? So you got on the line width the spectrum of wavelengths, VCSELs are narrower line widths. They’re not as wide as microLEDs. And so the same chromatic dispersion argument applies here because now your narrower line width will take your signal further before it spreads out. It still spreads. Chromatic dispersion happens to every form of light. Everything, right? But it happens only at a longer distance because the line width is narrower. So it has more room to spread before bits start to overlap.
The Future of Micro Emitters
Austin: Gotcha, gotcha. So you have stimulated emission, a tighter line width, which means you can send information further and already you can send information faster than you could with LEDs. And yet, by the way, to your point, this is gallium arsenide, so an abundant supply, and it’s still a pluggable form factor. So right, we’re not talking about NPO or CPO. This is just a cable that you plug in on both ends.
Vik: With Credo, yes, but there is plenty of talk about using micro-VCSELs for NPO, as we saw at the Astera booth. So those kinds of things are in discussion and there were some nice things we saw at the show floor. But all of that is coming, yes, but even now, I think we are pretty early in the whole VCSEL for interconnecting because even Credo’s booth did not actually have a cable that was showing electrical data on micro-VCSELs. They had some microLED stuff, but so this switch over, I feel is very recent. And there’s a lot of work to be done and there’s going to be a lot more stuff. I don’t think that we are done with microLEDs or it’s dead or anything like that because that’s why even Credo has formulated it as a micro-emitter. That’s the broad technology. There are still use cases for microLEDs, and the industry is still exploring them. But yeah, essentially they launched a work stream inside OCP called the Short Reach Micro-emitter Optics. So it’s SRMO work stream. So the whole point of this is like a bunch of companies are getting together to talk about how to do short reach micro-emitter optics. So there’s a lot of thing, lot of stuff that’s going to happen. This is not by any way settled. But I think the industry has at least for the near term chosen a path of micro-VCSELs. It seems that way at least.
Austin: Okay, okay. So micro-emitters is broad for it could be micro-VCSELs, it could be microLEDs, but you’re saying in the near term, it’s going to be micro-VCSELs because of all the benefits we discussed. It can reach 50 meters. It’ll be pluggable first is the pragmatic thing, but the same technology can be used for near-package and for co-packaged and we’ll probably work our way there. I guess then lastly, what will we see microLEDs anywhere and if so, where will we see them?
Vik: MicroLEDs are still good for one fundamental reason that I think a lot of the discussion misses is that they’re extremely heat insensitive. Gallium nitride is what they call a high band gap semiconductor, which means that the operation of the light source is very insensitive to temperature and that’s important in some applications. I would think one of those applications is scale-in where you use optics within the tray. The reach is not going to be meters, right? Like a tray, the server tray is only this big, about a few feet. And within that is a very hot environment. It’s very difficult to replace anything that goes bad there. So you want a very high reliability source of light. I don’t know the reliability numbers per se, but gallium nitride is in general a much better suited technology for something like that.
Austin: Gotcha. So it has a high band gap, so even if it’s in a very harsh hot environment, like chip-to-chip or die-to-die, where you’re just in the server and it’s just kind of always hot, the heat isn’t going to cause electrons to have all this extra energy and jump the band gap and start to cause light to be emitted, right? The high band gap is like, oh, sure, there’s extra heat, but the molecules inside this particular diode are not it doesn’t cause the LED to turn on, so to speak.
Vik: Right, right. So that’s the whole thing. So it’s high band I have a whole article on what band gap actually means on the Substack as well. So go look at that if you want to understand.
Austin: We’ll put all that in the show notes, yeah.
Vik: Yeah, link it. And then we can, because yeah, what you’re explaining is right. If you have a high band gap, then heat cannot cause electrons to switch.
Austin: Got it. All right, man. Well, I think this was a good micro-explainer on micro-emitters from microLEDs to micro-VCSELs. And with that, any closing thoughts or should we just call it here?
Vik: No, this is good enough for this episode because there’s a lot more to come. We spoke to several exciting builders in this space and we would rather have them come on the podcast to talk about all these technologies. We spoke to people who are all willing to come on Semi Doped, thankfully, who are building the light stuff, the chip stuff, and so we have so much exciting content planned after OCP.
Austin: Yes, love it. So we gave you the introduction here and we will bring the experts on and let them take it one level deeper. So with that, people, thank you for listening to Semi Doped. Like it, subscribe it, share it with people. We appreciate all your comments and we’ll catch you next time.
