II-VI Inc

NASDAQ:IIVI

Good day and thank you for standing by. Welcome to the Coherent Corp. FY23 Fourth Quarter Earnings Call. At this time, all participants are in a listen-only mode. After the speakers’ presentation, there will be a question-and-answer session. [Operator Instructions] Please be advised today’s conference is being recorded.

I would now like to hand the conference over to your speaker today, Paul Silverstein. Please go ahead.

Thank you, Kevin, and good morning, everyone. Thank you for joining our fourth quarter fiscal 2023 earnings call. Today on the call, we have Chairman and CEO, Dr. Chuck Mattera; Chief Financial Officer, Mary Jane Raymond; Chief Strategy Officer and President of Materials Segment, Dr. Giovanni Barbarossa; and Laser Segment, President, Dr. Mark Sobey.

As a reminder, yesterday after the market closed, Coherent posted a shareholder letter along with an updated investor presentation. They can both be found in the Investor Relations section of our website.

Before I turn the call over to Chuck for his opening remarks, I want to call everyone’s attention to our shareholder letter and accompanying change in format of this morning’s call. The shareholder letter contains the traditional financial statements that were previously set forth in our earnings press releases, along with additional color around our operating performance, key trends and outlook. Given the additional disclosures in the letter, we plan to devote the bulk of this morning’s call to answering questions from the financial community. We’ve undertaken this change with the goal of providing greater insight and clarity for our quarterly earnings release. We welcome your feedback.

I also want to remind everyone on this call that we will refer to forward-looking statements, including all statements the Company will make about its future financial and operating performance, growth strategy and market outlook, and that actual results may differ materially from those contemplated by these forward-looking statements. Risk factors that could cause actual results and trends to differ materially are set forth in the shareholder letter and the annual and quarterly reports filed with the SEC. Coherent assumes no obligation to update any forward-looking statements which speak only as of their respective dates.

In addition, during this call, we may discuss both GAAP and non-GAAP financial measures. A reconciliation of GAAP to non-GAAP measures is included in the shareholder letter. Unless otherwise stated, all financial information referenced in this call will be non-GAAP. Our discussion will be limited to those non-GAAP financial measures that are reconciled in the shareholder letter. Today’s conference call will be available for webcast replay in the Investor Relations section of our website for one year.

With that, it is my pleasure to turn the call over to Chuck. Chuck, please go ahead.

Thank you, Paul. I hope those of you listening in have had the opportunity to read our new shareholder letter.

In the fourth quarter, the Coherent team did a good job executing in the midst of a challenging macroeconomic environment. Our revenue of $1.205 billion was above the high end of our guidance and non-GAAP EPS of $0.41 was toward the high end of our guidance. Operating cash flow was $182 million, which marked sequential and year-over-year improvement. We invested $93 million in capital equipment, and we retired $121 million of debt.

When I look back on fiscal year ‘23, legacy Coherent contributed to our resilient business model. In addition, our track record following our acquisition of Finisar once again speaks to our ability to successfully effect strategic acquisitions, and thereby create shareholder value. Two major highlights in fiscal year ‘23 were related to our acquisition of Finisar. We demonstrated our unique scale while generating nearly 20% of our FY23 revenues from just two customers, one in Communications and one in Electronics. These are good examples of the strength of our vertically integrated platforms which enable breakthrough solutions, and our differentiated ability to scale to meet sudden increases in market demand like those that we are now seeing in AI. And while we experienced a surge in orders in Q4 in Communications for AI, the macroeconomic uncertainty that affected some of the industrial and instrumentation businesses, slower-than-forecasted recovery in China, and a post-COVID deceleration in the Communications markets, drove the conservative fourth quarter order patterns for some of our customers’ legacy products.

Recently, some of those customers have taken actions, including reducing orders of legacy products in the face of lower demand and reducing their inventory levels while slowing their planned investments in CapEx. This setup presents the ongoing challenge of managing through a retooling in fiscal ‘24, and so we got busy during Q4 to align our costs with market reality. We view this temporary slowdown in demand as an opportunity to strengthen our foundations. We remain bullish about the future because many of our largest customers are also resetting their strategies and accelerating their investments in new products that depend on our innovations and our ability to manufacture at scale.

The largest opportunity in FY24 that we are addressing is for 800G Datacom transceivers for planned artificial intelligence and machine learning build-outs. That demand should help offset the anticipated declines in demand from our traditional data center and hyperscale customers in data communications in fiscal ‘24. In addition, we continue our review of strategic alternatives for our silicon carbide business, another one of our major growth opportunities. Thanks to our strategy of diversification, we believe that we are well positioned to benefit from any improvement in the macroeconomic environment, though our outlook assumes that we will not see meaningful signs of recovery before the end of fiscal ‘24.

So, in short, we are prepared for a reset year and we consider these challenges as a temporary interruption of otherwise powerful secular trends.

Our guidance for the first quarter of fiscal ‘24 is revenue of approximately $1 billion to $1.1 billion and non-GAAP EPS of approximately $0.05 to $0.20 on 153 million shares. Regarding full year fiscal ‘24 guidance, revenue of approximately $4.5 billion to $4.7 billion and non-GAAP EPS of approximately $1 to $1.50 on 153 million shares.

With that, I’ll turn the call back over to Paul.

Kevin, if you could open it up for questions. Thank you.

Thank you. [Operator Instructions] Our first question comes from Samik Chatterjee with JPMorgan.

Hi. Good morning. Thanks for taking my questions. And thanks for all of the details in the shareholder letter, very useful to get all those details.

Maybe if I can start with just a sort of clarification question on the AI/ML, and particularly, sort of the inclusion in the guidance or deciding not to put it in the guide? Sort of -- maybe if you can sort of walk us through, are we sort of to interpret that you’re not putting any of those AI/ML sort of orders in the guide for fiscal ‘24, or is there sort of some amount of it in the guide related to what you have more capacity, visibility around? And you do mention sort of capacity ramp as one of the hurdles, I think, in the shareholder letter in relation to fiscal ‘24. So, maybe if you can walk us through what are you seeing in terms of capacity challenges? What you need to sort of see in terms of milestones to include that in the guide going forward? And I have a quick follow-up.

Okay. Thank you, Samik. I’ll take that. Maybe three points are helpful. The first one is that the Q4 bookings that we had a surge that we reported, that surge was all about AI. That’s number one. Number two, we have revenue for 800G transceivers contemplated inside our guidance in the $4.5 billion to $4.7 billion. There’s meaningful revenue for a delivery of 800G transceivers. Those have already started. In the first half of the year, we will see a ramp from Q1 to Q2, but the substantial amount of revenue we will deliver, it will be in the second half of the year. And what’s in front of that is managing our scale and especially managing our supply chain, and so we see the opportunity for over and above what we have in our plan. But that opportunity will require quite a few more synchronizations, including in the supply chain.

As we work our way through in the next few months through the first and second quarter, we’ll have our eyes set as we work to compete for the greater opportunity that may come inside this fiscal year.

Okay. Got it. Thank you. And then you did outline in the presentation -- in the shareholder letter, like, three separate opportunities in AI/ML relative to EMLs, your DFB laser as well as VCSELs. Any sort of thoughts in terms of broader terms where you see the most likelihood of success within those three buckets, and where most of these orders are coming in right now that you’re seeing? In which bucket is that coming in? Thank you.

Okay. Well, I’ll start out and then I’ll ask Giovanni to finish it off, Samik. As you know, we’re a vertically integrated supplier. And in addition to selling both VCSEL-based and EML-based transceivers for this application, we are also a supplier to the merchant market. And having said that, our ability to scale both at high performance and high volume and high quality to be able to meet this ramp is partly the basis on which we’re going to continue to win both new orders and to be able to ship.

Giovanni, please elaborate.

So Samik, I would say that roughly like, would be like one-third short reach, short wavelength, two-third long reach, long wavelength and -- which obviously, we can support both with our internal devices. And we see that kind of ratio being different in terms of volumes because price is different, so the volume will be probably 50-50. And that kind of ratio will probably be the same for the next five years, and I’m specifically talking about AI here, right? So the 800G, maybe even the -- in the future, the 1.6T, et cetera. So that kind of ratio will remain in the next few years, too.

Our next question comes from Simon Leopold with Raymond James.

Great. I appreciate you taking the question and providing all the detail last night, gave us a little bit of time to try to digest this. So, maybe a couple of aspects around the exclusion of the AI-related Datacom from the guidance. Could you help us understand the rationale for backing the AI-related business, which you described as worth several hundred millions, out of the guidance? And help us understand as well, if we wanted to include this in our own estimates, what do you think the impact would be on the EPS? Just as a very, very quick follow-up to this question, what assumption do you have in terms of your market share of 800 gig and above transceivers? Thank you.

Okay, Simon. Simon, let me clarify and let me repeat what I said in response to Samik’s question. Our guidance -- our revenue guidance for the full year and for the first quarter, but especially for the full year, contemplates a meaningful amount of revenue for shipping 800G transceivers. I want to repeat that. Our plan contemplates a substantial amount of shipments. I believe that there may be additional upside. That additional upside might be as much as $200 million. But for us to have the confidence to add it into our plan, we need to manage quite a few aspects of our ramp, and we’re going to go for it. But I cannot be sure that we’ll be successful in having everything come together in time to be able to have the confidence to add that $200 million or up to $200 million, but we’re focused on it. And if we have anything more to say 90 days from now, we will.

With regard to your second question, our belief is that we’re in a market leadership position. It’s a competitive market. There are strengths that we bring. And we’re going to continue to compete on the basis of the strengths that we have in this generation. Those strengths I outlined were evident in FY23 with a rather substantial ramp in Datacom transceivers year-over-year from ‘21 to ‘22 to ‘23. We have the ability to scale, and we have the laser components for the generations that exist today and the 1.6T, which are coming.

And so, on that basis, I think it’s the first inning. It’s very difficult to be assessing the -- what the score will be in the fifth inning and the sixth inning of the game. Everybody is just getting started, and we’re well positioned to be able to move even further past our own aspirations for our leadership position in this market.

That’s very helpful. And just to sort of paraphrase it to make sure it’s crystal clear, that the exclusion is because of risk of ramping the production capacity shipping. It is not because you believe this is sort of a onetime flash in the pan kind of project, this is the beginning of a cycle. Correct?

Right. This is the first inning of the game. It’s the beginning of -- I’m not even sure it’s a cycle. It’s the beginning of a revolution. And there’s a lot more to come, we believe. And by the way, if 90 days from now, I can update you on the possibility of additional revenue that we can add to the guidance, if we’re able to operationalize it, there may be more to come because this is just getting started, and we think that in ‘24 rolling into ‘25 is going to continue to drive our growth.

Our next question comes from Meta Marshall with Morgan Stanley.

Great. Thanks. Maybe just outside of the transceiver business for now. You had noted that visibility increased during the quarter. I just wanted to get a sense of whether that visibility increased anywhere outside of Datacom transceivers? And then, just as you look at recovery of the business, either late in the second half of the fiscal year or into fiscal ‘25, just what segments are most likely to kind of recover first outside of Datacom? Thanks.

Well, the Telecom business itself given our position of our portfolio, the strength that we have with the Telecom customers, I’m expecting that the second half of the year to be better than the first half of the year, and that we’ll expect -- I do expect to see -- signs of recovery before the end of this calendar year. Signs of recovery will be followed through on new orders, and those new orders will be consistent and commensurate with both the launch of some new products, including our pluggable DCO modules. And the ramp of those products from a base that we believe we can grow our share meaningfully beginning in the second half of next year. So Telecom would be one.

I do think the industrial market is broad-based. We’re diversified. And I don’t want to say that we’re at the bottom, but I do believe that any meaningful increase in macroeconomic activity, we will begin to see it both in laser utilization, in our service business and the further adoption of laser technology, including for EV battery welding as just one example. Our silicon carbide substrate business is growing, is growing super nicely, I would say, and I’m expecting that to continue into ‘24.

Maybe I’d stop. Meta, take a follow-up, if you like.

Yes. No. Just as a follow-up, I mean, just to come back to Datacom transceivers, when you talk about this kind of additional $200 million, is the majority of the gating item your capacity, or are there other gating items to kind of recognition of that?

Meta, every manufacturing line has a first constraint. And when it’s broken, there’s one right behind it, the second constraint. Our job and our expertise is being able to figure out how to manage multiple constraints at one time. Managing the supply line, we have critical components that we’re dependent on, and that supply line for the last six months or so has been in a wind-down mode with the tide going up. And the rather sudden adoption in the beginning of this -- the game in this first inning caught many in the industry by surprise. And therefore, we have to manage through certain aspects of the supply line. It’s going, but if it were going faster, we’d be able to take on more because we have the capacity to do more, and we’re aiming to do more. I believe that customers want us to do more. And so, in the next couple of months, the urgent approach to managing our entire manufacturing capacity will continue, and we’ll give an update in 90 days.

Our next question comes from Vivek Arya with Bank of America.

Thanks. One more on Datacom. For fiscal ‘23, what was your total Datacom transceiver sales, and how much was that in AI/ML? And I’m curious, how are you drawing that line. Is it anything about 200 gig? And so, that’s my first question.

Let me give it to you in broad strokes because I think that’s -- that will do it for you.

Our Datacom transceiver sales in ‘23 were approximately 20% of our consolidated revenues. And as it relates to AI, whereas there may be some applications and sockets that are at data rates that are less than 800. For us, when we talk about it, we’re really talking about 800, and you’ll hear us talk about our road map for 1.6. But in FY23, it was not a material amount. We began shipping in the third quarter, and it’s going to step up rather substantially in ‘24.

And for my follow-up, I was hoping if you or Mary Jane could provide the bridge between the sales guidance and then the earnings guidance. What are you assuming for gross margins in fiscal ‘24 and the exit OpEx rate in fiscal ‘24?

Okay. Thank you, Vivek. Mary Jane?

So, I think with respect to the whole year, we’re widening our gross margin range to 37% to 42%. And at lower revenues similar to what we have in the guidance, given the importance of volume, the margins may not be at 40% every quarter, so that’s the first thing.

The second thing is, similarly, at the same level of revenue that we’re talking about for the full year guidance, the OpEx tends to be a little bit higher percentage of sales, not because the dollar value of the OpEx is going up but because the revenue was lower. So it’s probably in the neighborhood of about 22% or 23%, which is the high end of our range of 20% to 23%. So that’s the way we’re looking at it. I do think that as the year goes on, we will probably see the margins improve as the volume picks up, especially if some of the outlook that we have that we think covers somewhere between the next two to four quarters starts to recover in the back half of the year.

But Mary Jane, how do we reconcile -- I mean, you’re guiding sales down, I think, what, about 11% at the midpoint and earnings down almost 58% at the midpoint. So, what declines a lot more? Because gross margins, from what you are suggesting, sort of seem to be about where they are in fiscal ‘23, unless I’m getting that wrong. So what are those missing pieces that are driving? Is it share count? Like, what is making earnings decline so much faster?

Comparing to ‘23, I mean, the revenue alone comparing just to the -- say the 4,700 -- $4.7 billion, sorry, the revenue alone is $0.91. So, the revenue declining is a significant impact on the earnings, and then from there, the taxes make a difference. The dividends change somewhat. Even though we have the share count converting on the Series A, the in-kind dividends also start to go up because they’re capitalized. So, those are probably the major things, but the revenue being below ‘23 is the primary driver.

Our next question comes from Tom O’Malley with Barclays.

I just wanted to see what silicon carbide was as a percentage of total revenue in the June quarter. In the prepared remarks, you talked about the continued supply constraints just from not being able to scale, but you also said it grew nicely. Could you just give it for the June quarter?

Yes. Good morning, Tom. Just give us a second. Yes.

The whole of the Wide-Bandgap product line was about 6% of revenue.

Okay. Thank you. And then, I just wanted to ask just a technical question, maybe this is in Giovanni’s camp. So, in terms of AI connections that you’re seeing today, I thought it was interesting in your slide deck, you showed that the biggest growing opportunity between ‘23 and ‘28 is the silicon photonics portion where you have it going from like $800 million to $4.6 billion. Where are you guys playing in silicon photonics? And why is that growing so fast in that period of time? Thank you.

Well, there are solutions that from a power consumption and generally speaking, performance standpoint, are better suited for silicon photonics. You still need a laser for those solutions. And we see -- I would say there is a split -- silicon photonics is not in the slide, it’s not identified. It’s not split between short reach and long reach, short wavelength, let’s say, or what we call it short reach, long reach. So I would say that split is still one-third short reach, two-third long reach, and both of them are related to the AI ramp that we talked about. So, that’s what we are seeing.

So the growth -- and of course, there is no -- that’s not all of it. That’s growing fast, but it’s also -- there’s a VCSEL part, which is also growing very fast, and that’s unrelated to the silicon photonics part. But just with respect to silicon photonics, that’s the split between -- still all the AI, but it’s a split of, let’s say, one-third, two-third short reach, long reach.

Our next question comes from Jed Dorsheimer with William Blair.

Just as a follow-up to the previous on silicon carbide, am I looking at this correctly that if I look at the Wide-Bandgap and I strip out indium phosphide, the silicon carbide’s roughly 70% of that business? And -- or is there anything else that I need to be aware of in that business?

Jed, can you repeat your question? Which set of…

What I’m trying to get at is the growth in silicon carbide, previously, you’ve said that that was 3% of sales. And now, Wide-Bandgap is a total of 6%. Within Wide-Bandgap, I think the other major component is indium phosphide, but that seems to be relatively small of about 20% to 30%, so the vast majority is silicon carbide. Is there anything else that would be -- that would need to be removed to get back to that silicon carbide? Because what I’m trying to understand is the growth year-over-year of that business.

Okay, Jed. Jed, in our Wide-Bandgap electronics platform, the majority of the sales are for silicon carbide substrates. That’s -- there were no indium phosphide or gallium arsenide, there’s nothing of any other compound semiconductors. But we do have our ion implantation services business, which includes providing ion implant services for silicon carbide to customers who are operating fabs. And so, it is silicon carbide-based. And that’s the focus of our Wide-Bandgap electronics platform, and it is growing. It is outpacing the growth of the Company.

And it’s been 4% to 5% for the last several quarters, Jed.

Got it. Okay. Well -- and then, I guess, maybe just -- you mentioned that the -- in the shareholder letter that 40% of the CapEx was directed to this business unit. And then last quarter, I think you updated saying that you were looking at a strategic review of this business unit. It’s a lot of CapEx for something that’s relatively small, albeit growing. So just how should we think of that percentage in the ‘24 guide? Do you expect a linear growth, or are you expecting with this capital intensity in terms of expansion that that would be nonlinear?

No, I’m expecting that in FY24 and planning that about 40% to 50% of our capital will be invested in silicon carbide to fuel the growth.

Sure. What I was getting at though, Chuck, is so, in a business that you’re looking at a strategic review, which could take different forms, that’s a lot of CapEx that you’re putting into that business. Are you expecting nonlinear growth in that segment? So in other words, if you’re at 6% right now, in that forecast, is that expected to double? Is it -- that’s what I’m -- to justify the CapEx in that business unit?

Mary Jane, would you like to?

Well, certainly, I think as we have seen and talked about probably in prior quarters, as the mainline vehicles from kind of mainline car suppliers start to move to electric, we absolutely expect that this market will have an inflection point upwards, and I think we’re only seeing just the beginning of that. The growth was very strong in the fourth quarter, it was strong in the third quarter, and I think we do expect that to continue.

Obviously, customers -- it’s -- fortunately, this business is somewhat more rational than other parts of our business. But it’s a -- it’s important for the capacity to be there for people also to commit to being able to change over half of their entire car lines, if not more than that.

It’s Giovanni’s segment. Giovanni, would you like to add?

So Jed, just -- I’ll give you a number. FY23, FY24, we expect growth of at least 40%, four zero.

Our next question comes from Christopher Rolland with Susquehanna.

Yes. I was wondering if you guys might be able to talk about any kind of nontraditional engagements. Have you guys had any engagements for optical or lasers into things like AI systems or servers or cards? And perhaps if you could talk a little bit more broadly about the economics for AI just in your transceiver business, how margins compare to corporate, what the OpEx needs are to support that business, et cetera?

Chris, could you repeat the first part of your question? The nonfinancial part?

Yes. So, your competitor, for example, is starting to talk about custom AI designs that are going into systems, not just transceivers. And I was wondering if any hyperscalers have engaged you -- any AI-specific companies have engaged you in custom designs?

Giovanni, would you take that?

So, let me start with the -- our strength in the device technology. We currently should sell more devices than we actually use internally. So, we -- most of our competitors are actually our customers too, as you probably know. So because of the broad-based laser and receivers technology platforms that we have work both in short wavelength, short reach and long wavelength long reach. So we are engaged on some, let’s say, nonstandard designs with some end customers. But even those customizations at the end of the day, in terms of guaranteeing interoperability, they will have to be standardized at some point.

So, I don’t believe there is any difference in terms of the process of standardization that we’ve seen in the past 20 years in the Datacom book. Of course, as the people trying to improve performance, cost of ownership, et cetera, there are ways to customize the solution internally. I want to remind you that at the end of the day, all of these x100Gs are all 100G optical lanes, so all comes down to 100G anyway. And so some of them are parallel, some of them are different types of approaches, multiplex and so forth. But at the end of the day, it’s all 100G in the optical lane level, even if there may be 200G electrical lanes coming soon in the future, of course, 200G optical lanes coming soon. But today, that’s where we are.

So, there is some level of customization that I don’t think it changes the -- again, our ability to compete is still substantially better than, I would say, most of our competitors because we -- as you know, we are the most vertically integrated out there, which doesn’t only improve our ability to, as we were saying earlier, to have and support the demand that we see, but also the ability to differentiate at the transceiver internal design level. So, we are engaged from that perspective, some of these super customer -- call it, customized solution for some of the AI players.

Mary Jane, would you…

With respect to the margins. So, generally speaking, you’ll remember that the communications entire end market, the margins tend to be below the corporate average. Having said that, the higher data rate and typically more technologically complex products that deliver a greater value, tend to have margins that are above the average within communications to a pretty decent extent.

Great. Thank you very much. For my second question here, I know it’s hard to figure this out and inventory at your customers, and I know it’s going to take a few more quarters here to work through. But is there any way to kind of quantify what this inventory burn is? What you think normalization would have been at your customers? How much more would you have shipped if you’re shipping in line with demand? Any thoughts on that? And then, any thoughts on perhaps the linearity of how this inventory dynamic plays out? Thanks.

Okay, Chris. I won’t be able to speculate on what it could have been. It’d be too complex a function and have great uncertainty, I think. But 90 days ago, I said I thought that the moderation would persist at least until the end of this calendar year and maybe longer, and it’s still not possible to call it any better than the remarks I made earlier. And as it relates to Telecom as one -- as a kind of a primary market where we were -- we are affected by it, I am hopeful for sure and looking and engaged in discussions with customers. I believe that we won’t see a turn up before the second half, and I believe that we will begin to see some signs of it by then.

Our next question comes from Ananda Baruah with Loop Capital.

I guess two, if I could. In the shareholder letter, you guys also talk about sort of AI exposures being more than just AI transceivers. And I think there is mention made of AI active and passive components, high-speed lasers. And I was wondering if you could drill down on that. I think you gave some context at OFC [ph] on this as well, but would love to get context on that. Any updated way you’re thinking about that? And then I have a quick follow-up. Thanks.

Giovanni, maybe just talk about VCSELs, EMLs and optics.

Yes. So, Ananda, I guess, you are wondering if the -- I mean, obviously, the vast -- the larger share of the revenue upside that we see in the year due to the surge in demand, which was, by the way, was a surge because it was unexpected in terms of size and timing, but not necessarily unexpected for the market trend because we know it was going to come at some point, but it caught us a little bit by surprise. And fortunately, we had -- we do have the 800G platform ready. So, it’s more a question of logistics and supply chain and ramping up, so that’s one thing.

But in terms of the split, let’s say, let’s call it, between modules and devices, let’s say, transceivers and lasers, photodiodes and in some cases, even ICs. Obviously, the transceiver is 90% of the total. So, that’s kind of the split that we see in the -- let’s say, in the next 12 months in the fiscal year, in the fiscal ‘24. So that’s kind of roughly the ratio revenue-wise.

That’s helpful, Giovanni. I appreciate that. And I guess, as a follow-up, very -- thanks a lot for the detail on the AI transceiver TAM in the slide deck and the go-forward view. Any opinion on what -- like, where you ultimately sit share-wise in the various buckets, the way you’ve laid them out? I mean just sort of bigger picture. And do you think you’re in a position to gain share going forward as well? Any context around that would be helpful. Thanks.

Ananda, we are the largest transceiver maker in the marketplace. And as it relates to this 800G opportunity and more broadly AI, including the generations to come, our goal is to be the market leader. It’s early as a starting point, but even though it’s early, we’ve gotten busy. I believe that in ‘24, what we have baked into our plan is already a super exciting ramp, and I believe that there will be a possible upside of several hundreds of millions, even in ‘24, in the back half of ‘24. And I believe that our ability to address the market, to serve the market, to scale in the market is going to be critical for us to be able to win. Our goal is to be the market leader.

And Chuck, is there anything about how you’re sort of going to market with -- with the sort of technology portfolio that you have that you think could increase your advantage in the market 800G, 1.6T, et cetera, relative to where you are today already as the leader?

Yes, absolutely. Let me repeat. Our laser-based technology, both short wavelength and long wavelength, and the demands on that laser technology, including for 800 but especially for 1.6, will separate us even further from the other players in the marketplace because we are the only vertically integrated company that has a road map to support well beyond 800G. And I believe that those are among the very strong value propositions that we present to a customer.

Our next question comes from Ruben Roy with Stifel.

Chuck, if I can ask you to put a little bit of a finer point on the AI/ML transceiver momentum you’re seeing. How much do you characterize the relative strength today, at least in sort of optics going into InfiniBand AI training clusters versus Ethernet? And really, what I’m trying to get to is if you can give us a sense of timing and qualification cycles for Ethernet deployments, say, 800 gig and then eventually 1.6T would be helpful.

Okay. Ruben, I’ll ask Giovanni to address both.

So, there is -- we are absolutely agnostic to the -- ultimately, the switch architecture of the customers, right? So, if you’re talking about maybe the market where the demand is, maybe there is a higher demand from InfiniBand than Ethernet and the rest, right? So, but I -- just from a hardware standpoint, it makes absolutely no difference to us.

Okay, Giovanni. So, are you qualified then for some -- I think some of the cloud service providers are talking about moving to new Ethernet switches, 51.2 terabit, et cetera. Are you qualified as those switches move out at some point in 2024?

Yes, absolutely.

Okay. And then, I guess, one last question then for Chuck or Giovanni. You mentioned the 20% of consolidated revenue in fiscal ‘23 related to what you would consider AI/ML. That leaves a pretty big portion of sort of how I would consider legacy transceivers, 200 gig, maybe even lower. How do you think that plays out? Do you think the AI/ML stuff is going to ramp faster than legacy falls off or legacy sort of hangs in there? How are you considering that?

Ruben, let me clarify. My earlier comment was about 20% of our consolidated FY23 sales were in Datacom transceivers. So, let me ask if you have any question about that.

Okay. No. I messed that up, then. Thank you, Chuck.

I’m glad we’re talking. As we indicated in the shareholder letter, we had a 10% customer in the communications market that was in the data communications market. And that -- the demand for those legacy products in my prepared remarks, I alluded to the declining demand for certain legacy products from our customers. That demand in FY24 is going to roll off. That is our plan. As it rolls off, even faster than it rolls off, I’m expecting that inside the fiscal year ‘24 that we are able to replace it with 800G AI transceivers at a minimum, and we’re aiming to do more than that. Is that clear?

Absolutely. That’s what -- so it’s contemplated. I appreciate the detail, Chuck. Thank you.

You’re welcome, Ruben.

Our next question comes from Dave Kang with B. Riley.

My first question is regarding your transceiver revenues. What was the split between different speeds, like 100 gig, 400 gig last year? And now, with 800 gig ramping, what do you think that mix will be this fiscal year? And what’s the margin differential between 100 gig versus 400 gig and 800 gig?

Giovanni, do you want to…?

I would say that the 200 gig and above was about maybe, I would say, 30% of the total, and the rest was 200 gig and below. However, let me tell you the split between the -- sorry, the other way around. Sorry…

So Giovanni, so 200 gig and above was 70%? Just wanted to make clear.

Yes. Yes.

Okay. Okay.

So now in terms of the major change in our revenue distribution has been -- if you recall in the past, we said that one-third was hyperscalers, one-third, let’s say, top 20 cloud and then one-third the let’s say enterprise and the rest. So as a result of the shift to AI and the result of some inventory in what we did over FY23, that ratio has actually changed now to two-third is around AI, at least for FY24, and that’s -- most of it will be 800G -- 200G and above, of course, including 800G. And then the rest is let’s say, smaller cloud players and enterprise. So that ratio, one-third, one-third, one-third is like two-third, and then the other one-third is split between the remaining cloud providers and the enterprise. So hyperscalers will be a much larger share of the total than in FY23.

And how does that transition, that shift impact the margins?

So as I noted earlier, the margins on the higher data rate, more technologically complex products tend to be higher than the average in a given group. So, in the case of transceivers, these greater than 200G and certainly 800G would be above the average. Having said that, for those parts that are, say, 100G and below, at some point would become very, very, very good at making those. So the lower margin tends to -- the lower-margin products or average margin products tend to hold actually because as time goes on, also, we’re almost the only one that makes them. So, generally speaking, that’s how you should think about the margin structure there.

Got it. And maybe, Mary Jane, just my last question is on your backlog, $2.7 billion. I was wondering if you can provide the split between your three divisions. And also, if you can provide -- I know you provided operating margins for three divisions. I’m wondering if we can get gross margins for those three.

So with respect to the backlog for the Materials segment, of the $2.7 billion, it’s about $650 million. Of the Networking segment, it’s about 1.2, and the balance would be in Lasers.

With respect to the gross margin of the segments, we don’t typically give that, as you know. But I think that you can imagine that it basically goes in somewhat parallel to the operating margin, remembering that grown materials and laser diodes have the highest margins in the company along with typically our industrial products. The laser systems are typically slightly above the corporate average, and the -- the communications margins tend to be below the corporate average.

The next question comes from Jim Ricchiuti with Needham & Company.

Mary Jane, I wonder if you could -- you’d be willing to share any targets for debt reduction. In fiscal ‘24, maybe, say, at the midpoint of your full year guidance, how should we be thinking about that? Then, I have a follow-up.

Well, the -- so we haven’t previously talked about targets for debt reduction. Obviously, we would have strived to be north of $200 million on that. Paying down the debt remains a very, very high priority, a very, very high priority, to the point that while the CapEx is the first call on the capital structure, we have been very, very aggressive about managing that CapEx because, to some extent, the CapEx naturally moderates in lower -- in periods of lower revenue.

And then looking at some of the comments you made in the shareholder letter on the display side of the business. I’m wondering if you -- are you expecting any kind of seasonal pickup in utilization that’s going to drive that part of the business in the first half. Just in light of what we’re hearing from one of the suppliers of OLED materials into the handset market. And then, on the systems side, you alluded to pick up in orders out of China. Is that deliveries for fiscal ‘24 or the bulk of those in fiscal ‘25? Thanks.

Great questions, Jim. Thanks for your questions. Mark, will you address both?

Thanks, Chuck. Hi, Jim. Yes, we definitely expect service utilization pick up in the first half of our fiscal year, so over the next six months. As you know, we get a batch of new smartphone releases from various manufacturers typically in the September, October period. And we’ve certainly got visibility into having expectations that that would drive our service revenue. And the bookings that we mentioned on our new OLED capital equipment, and the shipments are in FY24. So, those shipments are within the fiscal year.

Okay. But Mark, how does that compare with past investment cycles you’ve seen in this part of the business?

That’s a great question. I think it’s pretty similar. I mean, we’ve – we’d love this business to be more linear. It’s not. It tends to be associated with fab build-outs as you know, and we’ve got a limited number of customers, 6 to 8 major customers split between -- today, it’s pretty much split between China and Korea. So, I think it’s pretty similar. It tends to go in phases. Each phase is typically three to four systems, so you can imagine that the orders that we recently collected were in that sort of range. And we would expect additional phases to be built out as well as we’ve mentioned many times in prior calls, an expectation of Generation 8 fab build-outs, especially in China. So yes, we see this as reasonably typical.

Our next question comes from Tim Savageaux with Northland Capital Markets.

A couple of questions here and maybe they’re related, which is, specifically, I wanted to focus on trends in the Telecom business from a demand standpoint, and you probably have some inventory issues there as well. But, as you look through to end demand, what can you tell us about what’s happening there? And kind of related to that, if we look at the September quarter guide going down something on the order of $150 million, if you look across your segments, what are the main kind of drivers there in terms of puts and takes from a sequential standpoint? Thanks.

Okay. Thank you, Tim. We had very strong shipments in communications in the fourth quarter. I think that will be down into the first quarter. I think that’s the number one driver in the change from Q4 to Q1. That accounts for maybe more than half or two-thirds of the variation and the rest of it is spread across the other markets or across the other segments and markets.

What was the second part of your question?

I’m with you. So, to the point -- to the extent you’re talking about comps being half to two-thirds of the decline, I guess I was trying to get a sense of Telecom versus Datacom there.

It’s a combination, and the moderation that we saw in the March quarter in the Telecom market, that persisted in Q4 and it will leak into -- at least leak into Q1 as well. It’s a meaningful part of it.

Our next question comes from Mike Genovese with Rosenblatt Securities.

Just a couple of clarifications. First, the second customer, the one that was almost 10% the Datacom customer. I just want to clarify that. That sounds like a hyperscaler, [ph] and I just want to ask that. And then the second clarification is, when you talk about transceivers, you’re also -- you are also including cables, active optical cables and electrical cables as transceivers. I just want to check that nomenclature to make sure that cables are, in fact, transceivers in the way you guys talked about it. Thank you.

Giovanni, why don’t you take it?

Hey Mike, thanks for the questions. Yes, absolutely. Yes, we always include cables, but just to be clear, we have no electrical cables in our portfolio. So, it’s all optical. But we do include cables, absolutely.

And I think your first question probably was -- we couldn’t hear you very well, but I believe you wanted to -- yes, the two-third -- when I mentioned two-third, I was referring to hyperscalers. I think that was your question. So, we saw…

Sorry, if I could clarify that question. You said you had two large customers where orders were expected to go down and one is, obviously, the consumer electronics customer. I just wanted to check the second one, the Datacom customer, that that is, in fact, a hyperscaler?

Well, one -- electronics was really consumer electronics and the communication was Datacom. It was large customer of ours.

Okay. And then my next question, just one quick question. Just the 200G -- lasers, when do we expect those to be shipping revenue within your transceivers, or to external customers?

Well, right now, we are prioritizing our internal customer for anything 200G. And -- but we have design wins with some of our competitors, transceiver competitors. And so, we’ll -- that will be meaningful only in the second half of the fiscal year.

One moment for our next question.

Kevin, we’re going to wrap the call.

Okay. Ladies and gentlemen, this does conclude today’s presentation. We thank you for your participation. You may now disconnect, and have a wonderful day.

Thank you.

Thank you.