Lumentum Holdings Inc. (LITE) June 3, 2026 Conference Transcript

Welcome back to this afternoon session at the BFA Global Tech Conference. I'm Vivek Arya from BFA's Semiconductor Semicap Equipment Research Team. Really delighted and honored to have Wupen Nguyen, the head of business unit at Lumentum, join us this afternoon. I'll go through my questions, but please feel free to raise your hand if you would like to bring something up during the discussion. Very bold welcome. Thank you. I really appreciate you being with us. So maybe let's start at the high level, if it's okay. You know, I see Lumentum kind of involved in so many, you know, interesting high-growth areas, right, whether it is lasers, right, whether it is, you know, optical circuit switches or, you know, transceivers, right, just silicon photonics, right, multiple different areas. What are the key kind of operational and technical areas that you are spending your time on to make sure, you know, because there is no lack of demand, there's no lack of growth, but how are you kind of optimizing your, what are the top two or three operational things that you are keeping your eye on?

Right. So first of all, it's the fab, right? I think the fab is the most critical part of it. You know, our value is in the fab, you know, primarily. So we're really making sure that, you know, we have the right teams, right? One of the things that we are doing today with Lumentum, I'm seeing Chuck here, these FAB are actually like a temple. There's a lot of very experienced people running the temple. So we actually have to make sure that all FAB people are fully supported with the right tools, right automation to ensure they can actually maximize their efficiency. So number one is the FAB. We spend a lot of effort on the FAB. Second thing really is thinking about ramping all these kind of assembly and test operations is not easy, right? OCS is a complex product, and the modules is rapid cycle products. These ramping are challenging, right? So what we do here really is say, let's put our best resources, best engineering talent on ramping these complicated products. And frankly, we're actually moving our legacy products out of our own factories, moving to the CMs, for the CMs to help us manage those kind of products, right? And then the third piece really is supply chain, right? It's now AI demand going through the roof, everything from PCB to, you know, Garnet to, you know, lasers and PDs and DSPs, everything kind of short. So we're also spending a lot of time managing the supply chain piece, make sure that, you know, this do not become the bottleneck. Really the fab, the test and assembly piece, and also the supply chain piece. All right.

Makes sense. Nini, you have described almost kind of a 30% imbalance, right, between supply and demand on your, you know, EML, you know, lasers. When do you think that will close? Like, what needs to happen? Is there a certain time frame at which you see that closing? And as you look outside Wupen, do you see competition catching up and filling some of that gap?

Yeah, we're seeing now. So let's think about where we are in the AI space. Today, all of our so-called supply-demand imbalance is really still in a scale-out phase. And the scale-out is already taking all of us, all of our wherewithal, and we're still short. Right. And what we're seeing also is that, you know, optical scale up is happening in 2028. When optical scale up happens, the optical intensity increased by at least 3x, if not more. Right. And frankly speaking, if you look at, you know, we're struggling as an industry to even deal with the scale out demand for the next couple of years. And then before you can really fully, You know, get your head above the water, the scale-up demand actually happens, right? So frankly, you know, we're not, you know, too concerned about there's more capacity, you know, more components joining the game. We're actually focused a lot more on ramping our own capacity, right? Because what we see is that the CPU demand is as strong and on time or even stronger as before. We see the scale out EML demand, you know, stronger or equal or stronger than it's before. There's more demand for CW lasers. And then we're seeing this tsunami of scale up happening in a couple of years, right? So we're not overly concerned, frankly, by competition, but we're actually much more focused on ramping our own capacity and driving as much efficiency as we can.

Got it. You made an interesting point that, you know, the specifications of these lasers, right? I mean, it's a similar laser, but, or same technology, right, for these EMLs can vary as you go from scale out to scale up. Could you expand on that? Like, what has to change?

Yeah, so I think there's also a very important point on the competition, I want to emphasize a little bit, right? Not every laser is built the same. Not everyone is the same, right? For example, right? Let's take the UHP laser or CPO laser, for example, right? It has to go through a very special modulator, signal photonics, in order to get the density of the CPO that customers are looking for. And that requires a very narrow line width of the laser, very low noise. That's hard to make, right? And then that, as you yourself, is a very high bar to overcome. And then for the CW laser, for example, when you go to a high-efficient CW laser, you actually are sacrificing certain features of the laser from a design point of view that makes yielding the laser a lot harder. And therefore, we are a user of different CW lasers, too. We see a drastic difference between brand A to brand B to our own. Right? That the design of the laser, ability of a laser to yield in manufacturing at the module level is a really big deal that people do not talk about. Right? So that's also really important. There's a major differentiation there. And this is frankly where Lumenta's pricing power actually exists because our customers are telling us when they use our lasers, they actually get a higher module yield. Now at the module level of a few hundred dollars per module, a couple of percentage difference in yield is a lot of money. So that's also really important, right? Third, at the EML situation, right? When you go to 100G, 200G EML, it becomes harder to make. And then you also have to go to the structures that will reduce the interference between different channels. That requires a different feature on the EML, right? That also makes it harder to make, right? So as we progress toward scale out and then scale up, the challenges, because of higher density of packaging and then higher density of bandwidth, makes the design features that much more important. And then the yield consideration not much more important. And so therefore, not all lasers are created equal at the final solution level.

Got it. That's a good point. Does that show up in pricing? Because people have always thought of the optics industry as, you know, kind of even more deflationary than the broader semiconductor industry. Do you think it is helping from a pricing perspective also?

Yeah, we hate to say this, right? But, you know, in the tele... What do you mean you like to say this? We like to say that we're over that face, right? Previously, you know, I've been in the industry for 30 years, and, you know, Chuck has been there for 30 years, too. I mean, that was a kind of zero-sum game market, right? Today, the market demand is growing so fast, right? If you have a good quality, you can supply. The pricing power does exist, right? So we are seeing, you know, certainly there's a structural level pricing. We're also seeing opportunities for upcharging when there's a last minute hot demand that when competitors, you know, fail to deliver. And we'll see opportunity there for us to also, you know, upcharge based on the local supply and demand mismatch as well.

I wanted to, you know, get your perspective open on the rollout of CPO, right? So maybe first start with what is kind of the most important milestones that you're looking at? How will it evolve? And then the question that I had, right, following that would be, you know, there's a lot of talk of CPO from the suppliers, not yet from hyperscalers quite yet. But maybe first start with, you know, how do you see the timeline of CPO deployment across your customers? Okay.

So first of all, you know, on the CPO side, right, we're seeing our ramp of CPO lasers to be on schedule. And then the pressure from the customers are the same or increasing, right? That means that they are all in on the CPOs. In fact, we heard some of the direct customers, the NeoClouds, are saying that definitely, right, CPO is on schedule. But I want to kind of bring the stage out a little bit. What is driving CPO? Why is it very important, right? The CPO is never a scale-out story, right? The CPO really is a scale-up story, right? If you look at the AI hardware racks, the scale-up capacity or bandwidth is 10 times that of optical scale-out. When you go to 10 times the bandwidth, it's hard to say, I'm going to use pluggable modules. The power consumption, the cost, the size is just crazy. And so that's one thing, right? And because of the inference, the AI inference traffic now growing so fast, I just made a statement that we're seeing now the inference-induced optical scale up, you know, ramping happening in Canada 28, right? So to meet that timeline, that's what's fundamentally driving the CPO, because you need the density of CPO, the power consumption of the CPO, to really enable the inference optical scale-up kind of timeline, right? So that really fundamentally is driving it. Therefore, we're seeing more urgency in the CPO adoption. We're seeing more urgency on the high-density optical interconnect adoption for that particular reason. And therefore, again, from the laser supplier perspective, we see that trend continues. So for the next couple of years, we're still, it's mostly in a scale-out domain. But in 28 and beyond, once the scale up happens, there's inflection of the optical intensity, right, in the AI racks. And that's going to be another huge opportunity that we've been talking about, right, since the OFC time.

Got it. Do you think, does that transition leave a hole that, you know, because CPU is implemented in scale out first, that, but that's a very temporary, you know, part, does that then leave a hole that scale up then has to fill? Or do you think, as you described, scale up is such a massively larger opportunity that it shouldn't be a problem?

Yeah, I think the eventual state with the following, right? The scale up is a very proprietary setup, right? You know, hypercellular A to B to C or, you know, GPU A to B to C, they have a different kind of design for the rack level, right? And therefore, scale up CPO is very straightforward. It's my own design. I own the racks. I do CPO, right? But the scale out is different. Scale out, you're connecting different racks together. That's where you may not have to use. CPO. You could use DR optics, FR optics, LR optics, right? That's where you can use pluggable. It's okay, right? So again, I would say that the scale out CPO is the first step, is to really trailblaze and mature the ecosystem, right? But that's not the final destination. The real reason for scale up for CPO is in the scale up. And you will see, I think, a mixture of optics used for scale out, including CPO. But you're going to see a lot of CPOs in the scale-up applications because they're still contained. That's how we're seeing the world developing.

Where does NPO fit in this? Is that a temporary solution? Does it have legs? What do you think about it?

Yeah, NPO is a great question, right? So here's our view, right? Still coming back to the scale-up story, right? When people realize that, you know, scale-up has to happen because of the inference impact, right? People figured out that, you know, our view is that they're not ready for CPO yet, but you cannot use pluggable, front panel pluggable. What do you do? Right? Okay, well, let's use an NPO, which is a socket-based, which is more like linear optics, close enough, not quite as good as the CPO internal power consumption, but it's much better than front panel pluggable, right? So it is really an intermediate first step to really for time-to-market reasons. But what we're seeing as a general conclusion is that the next generation after the NPO will be CPO. Everybody is focusing on CPO.

It's just a temporary thing.

So we really think, OK, there's going to be two group of people in the first generation of the scale-up implementation. It will be CPO plus NPO. But in the next phase, right, the generation two, everybody will be focused on a CPO. So this is really, in our opinion, NPO is intermediate step.

Got it. Now on the CPO, the question I really wanted to ask was, so I think it's clear NVIDIA has described the roadmap, right? They have kind of very clear timeline of when they want to get to CPO. But if you're not deploying the NVIDIA solution, right, then are you still seeing enough of a pull to, so yes, I think there is a, you know, Marvell, you know, Celestial, right, solution with Amazon as an example. But are those really kind of the two alternatives or do you see a more diverse set of customers? Because we don't hear yet from the hyperscalers, you know, because this is brand new technology, right, they're already putting in so much complexity. So what do you think? What do you hear from the end customers about the pace of adoption? What are their practical challenges that need to be addressed before CPU becomes a bigger deal for them?

That's a very complicated question. Every hyperscaler is very different, right? One hyperscaler is very much because the unique architecture, right, is very much a pluggable play, right? It is OCS, pluggable. That's one group, right? That customer, you know, they're probably not so hot on CPU, right? But then there's definitely NVIDIA enables every single hyperscaler to use CPO. So that's one thing. And then we also see another hyperscaler that's all in on NPO, right? That's, again, because optical scaled up, right? And we see a couple more customers who are more looking at, OK, I want to go to CPO, but I'm going to use a merchant solution first, right? So there's a merchant solution on CPO, including the switch. That's also where it's going to go with. So I would say that for the next two years, two to three years, you're going to see the pluggable customer with their own architecture. You're going to see the NVIDIA with the CPO architecture. And you're going to see some other hyperscalers, say, hey, whether it's a high density XPOs or whatever, or solutions, it's going to be a mixed bag, right? But that will change. Again, after 2028, it will change, right? To be more on, you see the, sorry, I'm a little bit over, you know, look at this OCI MSA. Who's on the OCI MSA? Every single hyperscaler on the OCI MSA, right? The OCMSA has to be done by silicon photonics, by ring modulators, by slow and wide optics, DWDM, right? That is the ultimate solution, right? So that's where everything is going to converge. It's just everybody has a different starting point. We're going to actually gradually converge onto CPO from kind of different angles. But the final solution is pretty clear.

Does it make your job of forecasting tougher, or you think that, you know, yes, all these things sound very fragmented, but you have very good, you know, line of sight, good visibility, right, for each of these implementations that you can roll up, right, for your factory forecasts?

Yeah, I think we have a good visibility, right? We have a good view on which hyperscale is doing what, right? And we're engaged in, you know, on the laser piece, we actually engaged very well, right? In all different opportunities. We also engage on the module side, right? In some of the major application of modules, right? So I feel we have a really good understanding. And also there's like multiple LTAs, right? Well, we have, you know, with the customers that also give us a visibility of the product types, product mix. Right. And the volume. Right. So I think we're pretty well positioned. That's why we're pretty aggressive in ramping our capacity. Right. Because we have these kind of visibilities and in most cases, the commercial assurances as well.

Got it. Makes sense. Next thing on optical circuit switch. Again, from the outside, it seems like, you know, there is one big hyperscaler who has adopted it, but, you know, Lumentum has described engagement with several hyperscalers. So first, the use case, is this still kind of limited more to, you know, high density kind of training clusters? Or do you see the role of OCS expanding in inference? Also, how is it doing that? And then are you seeing engagements with multiple customers? Or is it still kind of, you know, dominated by one?

That's a really good question, right? So today, we have been engaged with two kind of opportunities. One is more of a spine level replacement, whether it's really data center spine or, you know, campus level spine or scale across spine, right? It's really got spine level replacement. That's a smaller opportunity, right? But the primary opportunity here really is the optical scale up, right? In the particular Taurus kind of implementation, right? That's the majority of the volume, right? Then, what we see actually going forward is that when the CPU or optical scale-up happens, then there is an interesting phenomenon we're thinking really is that that's the time when you are scaling the inference clusters from 1 to 8 or 10 or whatever, go to 1,000 XPU clusters. That's where, you know, OCS becomes a really interesting opportunity, because then you can have different instances, you can have different models, you can actually serve, you can connect and then should leverage the whole clusters and not wasting any resources, right? That's when we see that the OCS group will become a really good complementary technologies to go with optical scale up, right? So we think there is an interesting potential convergence of optical scale-up and OCS. Not right away. The world takes time to really learn how to operate in that domain. But I think there's a multi-year trend that this could become another major growth opportunity for OCS on top of the current application and on top of the scale-across and scale-out applications of OCS.

Got it. So I know with CPU, right, you have a major system developer, right, really kind of promoting it. Do you see some kind of similar trend? So yes, OCS has been deployed by one hyperscaler, but for it to be widely deployed, do you think some system vendor has to really start pushing it? Or do you see the organic pull from end customers to deploy it also?

That's a good question. I think the, I would say that now I think it's pretty clear that, you know, the inference requires clusters, racks, right, racks and clusters. And we see hyperscalers thinking about how they can build their own clusters, right, really for high efficiency inferencing. I think the XPU guys, they're all thinking about it. Now, to learn how to use it will take time. Google spent many years perfecting the use of OCS. The software piece is pretty complicated. Therefore, I think that whoever wants to use OCS going forward needs to go through that learning cycle. I don't think it's going to be overnight kind of thing. But again, once the signal goes to optical in the scale-up domain, it makes sense to process in the scale-up domain. That's right. Why make the OE conversion? Why go back? I think that's a natural kind of migration for the world to go in that direction.

Got it. That's a very good point. Last question on OCS. Do you think there is still industry debate on the specific technology, MEMS versus liquid crystal? I imagine you have your own perspective, but do you think that debate has been sorted out, both technologies work well, or do you think MEMS has a better shot at being more successful?

No, I'm highly biased, right? Let me just say that, right? I'm highly biased. But if I look at just a time to pull outside, look at technology, right? The way I look at it is that the MEMS is an engine you build in parallel. You build at the wafer level. You build an engine, and you put optics around it, you put a fiber alignment to it, and you're done, right? The liquid crystal is more of a serial approach. You have those two layers and layers and layers of liquid crystal, right? So it becomes more a serialized manufacturing process, and thereby you introduce a little more insulation loss and so on and so forth, right? So I think manufacturing process itself is different and resulting in potentially different performance optically or potentially different manufacturability, right? But we don't have liquid crystals we don't really know. And third piece, really, some people are using the piezoelectric, right? That's probably even more serialized, right? And therefore, our view is that we know MEMS is the best. We know it meets all the specification. We know we can manufacture it, right? We know we're ramping it. So I think jury is out, right? I think if the market is big enough, it needs more than one supplier that can do this, right? Again, if the whole market goes to the inference-based, optical, scaled-up domain big enough, the market will need multiple technologies to support it. But I think we know our technology is pretty well suited for the application, and we can actually scale it.

Next topic, Wupin, I would like to move to is optical transceivers, like a bulls. Right. Lumentum, I find, has been a little reluctant initially to invest too much into the business. Then you got CloudLight, and then you got an important customer. So how strategic of a market is optical transceivers for Lumentum? Is it a lot more strategic, or is it still more tactical that you have one or two customers, you will do what they need, but you don't want to make this a big business?

Yeah, I think Optical Transceiver is an interesting piece, right? I think the, so first of all, there's a lot of questions in the conference saying, hey, when CPO happens, what's going to happen to your module business? And luckily, I would say that our primary constituent of the module business was customers who don't like to use CPO. So there's no kind of motivation, right? We're kind of enjoying the benefit of module ramping along with, you know, OCS ramping along with, you know, the CPO ramping. So we're kind of lucky in that sense. But, you know, the module piece is pretty important to us because through modules, we actually get to see a lot of the roadmap piece for our customers, right? Whether it's high density, like XPO or next generation architectures, where things are moving, right? And then there's a lot of evolution happening now, technology space around the modules. So I would say that's strategic in that way, but we do admit, right, the margin is definitely below our corporate gross margin, right? So our goal is, let's serve our current customer well. Let's not go crazy to expand our customer base, but let's just focus on serving our couple of hyperscaler customers really well, and then right along with their roadmap, enabling them, and continue to develop the technology. So I would say it's strategic, but it is kind of a confined strategic product line for us. Got it.

Is there a scenario where, because of various geopolitical reasons, U.S. hyperscalers decide to buy more design in the U.S.? I mean, everyone is kind of making it in Asia, so I don't know whether that makes a difference or at least packaging it there. But do you think there's a scenario where you get some advantage and that might make Transceiver a bigger business?

That's a really good question. I think that there is a general concern and hesitation on the hyperscalers of the dependence on the China-based module vendors, even though they're all moving the production outside, right? Now, different hyperscalers have a different level of concerns. Some are very concerned, some are less concerned, right? Very hard to say what they're going to do, right? Because frankly speaking, the China-based vendors are doing a really good job, right? Serving the needs of the hyperscalers. Without them, probably our AI will be, you know, it's not going to be that well either, right? So I think it's a complicated question, but we do see concerns. And we do see concerns about the IPs, manufacturing location, where things are coming from, and so on and so forth. So I cannot predict the future, but I think, you know, the concerns do exist. Yeah, but again, the AI, you know...

Have customers spoken to you about, you know, like, let's say if they wanted to, for whatever reason, do you think you have the desire, the capability, the capacity to actually become a bigger transceiver vendor?

Well, we don't have the capacity for sure.

Okay, well, that takes care of it.

But you would, you know, given time, anything happened, right? You know, given two years, can you build capacity? Yeah, you can build it in two years, right? Right. And it depends on what is the commercial condition, right? Would that be worth our capital allocation to make that a much, much bigger business? Technology, we have it. We have some key components. But whether we do that or not, depending on the commercial conditions, I think, by that time.

Makes sense. And then finally, on long-term agreements, you know, it seems like this year everything is in short supply, right? So customers, I imagine, are willing to, you know, kind of sign up for whatever, right? But how durable, how enforceable, right? How are these, obviously not, you know, proprietary agreements, but how in general are you structuring them? Are you holding customers to a certain unit commitment? Are you holding them to a certain pricing commitment, right? Percentage of volume commitment? What is the general nature of these LTAs?

The general nature of that is basically a take-or-pay kind of thing, right? So it's very take-or-pay. There's a very strong commitment from the customer side, and in some cases, there's a prepaid element as well, right? So it's very much structured like that, right? Again, And the duration depends. Typically, you know, two or three years or something actually longer, right? So it definitely is also, we're also thinking strategically. We don't want to, you know, max our capacity with LTAs because we still do believe that the AI optics, you know, usage will continue to rise. Therefore, we don't want to just lock up our capacity we have. We want to preserve some capacity for engaging more customers or, you know, opportunistically spot price of the market and so on and so forth, right? So that's kind of our overall strategy. But by and large, it's a take or pay kind of architecture of kind of the framework of the LTAs.

Got it. And how long do these run till, some of these LTAs?

No, it's like, you know, three years to five years. Two to three or even five years in some cases.

If, let's say, theoretically, cloud capex slowed, for whatever reason in the next two years, you feel that these are enforceable?

We feel it's enforceable, but are we going to sue a hyperscaler? Who knows? But I think that these are important relationships, because I don't believe AI is going to crash. You may slow down, but you're going to grow again. And the reason we're building with hyperscalers are truly strategic. And therefore, you know, I hope that, you know, they view it that way too. But we don't see that being a transactional relationship.

Okay. Last question, you know, NVIDIA made an investment in Lumentum. So from what has been kind of, you know, said publicly, what is the nature of that relationship? You know, what kind of assurances, what kind of engagements do you have because of that relationship? What does Lumentum need to do, right? What is your part of the relationship?

Yeah, that's a great question, right? So basically, there are two pieces, right? The piece number one is the purchase agreement contract, right? That's a multi-year LTA for the ultra-high-power laser, right? That is actually separate from the investment piece, right? Investment piece is completely independent from the purchase agreement, right? So investment really is like to signify the strategic relationship between the two companies, right? And then, you know, that's an important thing because then both banks will say, okay, we are strategic partners, but there's no string attached to what we can use that funding for, right? They just become a shareholder of Lumentum, right? And then that, the also high power laser for CPO, right? That's something we're ramping to. There's a commitment on both sides, right? There's a volume, there's pricing, there's take or pay.

Will you be exclusive, you think, in that opportunity?

No, I don't think that's an exclusive relationship, right? I think also because we don't want to be exclusive. They don't want to be exclusive, right? So there's going to be at least two companies sharing that share, which I think is good for both companies. Great. Makes sense.

With that, Wilpin, thank you so much. Really appreciate your time. Thank you very much.

Thank you.