Ciena (CIEN) Q4 2025 earnings call transcript

Ciena Corp (Q4 2025 Results) December 12, 2025 Corporate Speakers • Gregg Lampf; Ciena Corporation; Vice President, Investor Relations • Gary Smith; Ciena Corporation; President and Chief Executive Officer • Marc Graff; Ciena Corporation; Chief Financial Officer • Scott McFeely; Ciena Corporation; Executive Advisor Participants • Ruben Roy; Stifel; Analyst • Simon Leopold; Raymond James; Analyst • Atif Malik; Citi; Analyst • George Notter; Wolfe Research; Analyst • Tal Liani; Bank of America; Analyst • Samik Chatterjee; JP Morgan; Analyst • Timothy Long; Barclays; Analyst • Ryan Koontz; Needham; Analyst PRESENTATION Operator^ Good morning.

Gregg Lampf^ Thank you, [Dru]. Good morning. Welcome to Ciena's 2025 Fiscal Fourth Quarter and Year-End Results Conference Call. On the call today is Gary Smith, President and CEO; and Marc Graff, CFO. Scott McFeely, Executive Advisor, is also with us for Q&A. In addition to this call and the press release, we have posted to the Investors section of our website an accompanying investor presentation that reflects this discussion as well as certain highlighted items for the fiscal quarter and year-end. Our comments today speak to our recent performance, our view on current market dynamics and drivers of our business as well as a discussion of our financial outlook. Today's discussion includes certain adjusted or non-GAAP measures of Ciena's results of operations.

A reconciliation of these non-GAAP measures to our GAAP results is included in today's press release. Before turning the call over to Gary, I'll remind you that during this call we'll be making certain forward-looking statements. Such statements including our quarterly and annual guidance, commentary on market dynamics and discussion of our opportunities and strategy are based on current expectations, forecasts and assumptions regarding the company and its markets, which include risks and uncertainties that could cause actual results to differ materially from the statements discussed today. Assumptions relating to our outlook whether mentioned on this call are included in the investor presentation that we posted earlier today are an important part of such forward-looking statements, and we encourage you to consider them.

Our forward-looking statements should also be viewed in the context of the risk factors detailed in our most recent 10-Q and in our upcoming 10-K filing. Ciena assumes no obligation to update the information discussed in this conference call, whether as a result of new information, future events or otherwise. As always, we will allow for as much Q&A as possible today, though we ask that you limit yourselves to one question and one follow-up.

With that, I'll turn the call over to Gary.

Gary Smith^ Thanks, Gregg. And good morning, everyone. Today we reported record fiscal fourth quarter and full-year revenue of 1.35 billion and 4.77 billion, respectively. These records are a direct result of our sustained, purposeful investment and focus on leading high-speed connectivity technologies, together with disciplined execution and deep collaboration with our customers. Combined, these advantages have positioned and will continue to position Ciena to deliver value in the AI ecosystem, serving both cloud and service provider customers for many years to come. Our progress in driving value from our operating model is also reflected in Q4 earnings per share of $0.91, up 69% year-over-year and full year EPS of $2.64, up 45% from fiscal 2024. And lastly, we generated record orders for the year of 7.8 billion which resulted in our entering this year, again with record backlog.

These strong results really underscore our overall market leadership position, as well as the ramping broad-based demand across our business. And to this point, I'd like to provide some insights into what we believe to be robust and durable demand over the next several years. Firstly, we continue to see accelerating demand from our cloud customer providers. And that includes the large hyperscalers and the emerging neoscaler segment that we talked with you about last quarter. In fact, cloud providers today are as focused on scaling their network as they are on their access to power. Orders from cloud providers are very strong, and ramping across our portfolio and they constitute a substantial portion of our growing backlog. It's important to note that this accelerating demand is being driven by several dynamics.

And as counterintuitive as it may seem, the cloud providers largely actually underinvested in their networks to date, particularly relative to other areas of AI infrastructure. The major hyperscalers, who are seeing rapid traffic growth, not only have the capital to invest but also have real sustainable business models that are currently constrained by the need to dramatically scale their global networks. These cloud providers cannot and do not intend to strengthen their significant investments in AI-related data center infrastructure. And I would stress that that traffic needs to leave the data center to be monetized and operationalized, and we are their strategic technology partner for those network requirements. Secondly, demand from our service provider customers continues to grow steadily as they, too, reinvest in their transport infrastructure after years of digesting accumulated inventory, and also be focused on other areas of their networks, most notably and specifically 5G.

In addition, service providers' businesses are also being fueled by AI through the enterprise cloud demand, and specifically, cloud providers need for managed optical fiber networks or MOFN. And as a proof point here, we have recently won and are working to deploy a large MOFN project in India with two service providers for a major hyperscaler. Additionally, in the quarter, we have secured multiple major MOFN wins in other regions including several in new and emerging geographies for our business. And as a result of these dynamics, service provider orders were up nearly 70% for the year. And in fact, our top three service providers revenue from '24 to '25 grew 16%. Due to the increasing momentum across both cloud and service providers, Ciena's optical market share has continued to grow and extend our overall leadership, adding two points year-to-date, and we expect further gains clearly in 2026.

In order to address this accelerating demand, we are committed to increasing investments and working with our supply chain partners to scale the business. With product delivery lead times extending in the face of this unprecedented demand, we are proactively expanding our capacity to ensure our ability to timely meet our customers' demands. Indeed, this has already yielded results for fiscal '25 as we delivered double our initial revenue growth expectations for the year. Marc will discuss how we are stepping up investments to support demand and the expanding opportunities we expect over the coming years. The simple truth is that AI continues to drive network expansion across all our customer segments, and the scale of investment currently underway is massive and accelerating faster than anything we or, indeed, the industry have seen to date.

I would also mention that unlike the COVID-inspired supply-demand imbalance, we are seeing this demand be installed and leveraged for real near-term revenue opportunities at our customers, as evidenced by accelerated implementation services that increased in revenue by 34% in fiscal '25.

With that, I'd like to take a moment to share our sort of broader perspectives on the AI opportunity, as it relates to high-speed connectivity. As bandwidth continues to grow inside the data center and as this traffic flows out of the data center, AI inference models are moving closer to the network edge. And for the reasons that I mentioned earlier, we will continue to expand our existing leadership and addressable market in high-speed connectivity in the WAN. In addition to the wide area network, we're also seeing a significant addressable market opportunity in the -- in and around the data center. It is, I think, well understood that cloud providers are investing heavily in data centers to deliver on the current and future promises of AI.

Many third parties are estimating capital spending of more than $7 trillion through the end of the decade in all AI-related infrastructure. And this is, obviously, necessitating the need for both training and inference workloads at massive scale. As a result of the massive growth in AI workloads, and to address growing power and space constraints, cloud providers are planning and building distributed AI data center training clusters or AI factories which require multiple clusters to act as one. In fact, along with those power and space constraints, the ability of the cloud providers, and specifically, the major hyperscalers, to scale their global networks is becoming the critical long pole in the tent for them to operationalize AI for both training and inference purposes.

Within these data center environments, there are three key connectivity requirements, to scale up within a data center rack, to scale out between racks in a data center, and finally, to scale cross between geographically distributed data centers, which must operate at the highest levels of performance with super high capacity and the lowest latency possible. With our innovation and time-to-market leadership in high-speed connectivity solutions, our position could not be better to fulfill this critical demand. This growing AI-driven opportunity for Ciena is what we refer to as in and around the data center. In fact, our in and around the data center opportunities grew threefold from '24 to '25 and are a major contributor to our '26 expected growth rate.

We have proactively invested in our portfolio to intersect this growing market segment and with a few notable examples. First is our interconnects portfolio, comprising both our power and space savings ZR and ZR+ pluggables and our optical components. We expect Interconnect to play a meaningful role in scale up, scale out and, in fact, Scale Across workloads. In fiscal year '25, we surpassed our target of more than doubling FY'24 pluggable revenue, reaching revenue of more than $168 million. In the quarter, our WaveLogic 6 Nano 800-gig pluggables are shipped for initial revenue. And since the end of the quarter, we have shipped 800 ZR plugs to three additional cloud providers for testing and certification. And with regard to components, we address the cloud providers preferred disaggregated consumption model with our high-speed Coherent and other industry-leading WaveLogic technologies, including a DSP, Curtis, and other high-speed analog and electro optical components.

In addition, our components business now includes the electrical and optical interconnect solutions from our acquisition of Nubis Communications. The Nubis technologies and expertise will help us address the scale-up and scale-out opportunities inside the data center. We're excited to have the Nubis team as part of Ciena and are on track to GA first products in fiscal '26. And as we've previously noted, as technology advances and data rates increase, the components portion of our Interconnects portfolio primarily represents revenue opportunities beyond fiscal '26. In addition to our Interconnects portfolio, as market needs continue to evolve, driven by AI, we're seeing new architectural applications arise in and around the data center with two recent cloud provider use cases, I think, of particular note.

First use case is the Scale Across architecture, linking geographically dispersed AI training clusters using our market-leading RLS photonic line system, wave servers and Interconnects portfolio. This is an opportunity we discussed over the past couple of quarters, where a large hyperscale is linking to two regional data centers to build an AI backbone. I'm pleased to report that this hyperscaler is now extending this architecture to more locations. Additionally, I am pleased to announce that two more major hyperscalers have chosen our optical solutions for their Scale Across training applications as well. The second use case is out-of-band network management. Ciena's unique DCOM solution leverages our XGS-PON and other routing and switching products and was initially designed with Meta to meet hyperscale requirements.

Today, I'm pleased to announce that our DCOM business with Meta has expanded, as they plan to deploy in multiple new data centers. Also we're engaged in advanced technical discussions with additional hyperscalers to deploy this DCOM solution in their data centers. I'd like to briefly acknowledge here that the Scale Across and DCOM wins are just the most recent AI-related use cases to materialize for us in recent months. They are great examples of how we co-create and productize with market-leading solutions to address critical customer scaling requirements. And we fully anticipate continuing to develop additional innovative solutions with our customers, as they monetize AI across the various architectures.

Before I turn it over to Marc, I really want to reiterate that as we leave Q4, and indeed, the entirety of 2025, we have absolute conviction that the positive market dynamics and our technology leadership provides us with increasing confidence that the durability of demand and our business and financial trajectory are very strong. I'll now hand it over to Marc for a closer look at our Q4 and fiscal '25 performance and outlook. Marc?

Marc Graff^ Thank you, Gary. And thank you to everyone for joining the call this morning. Before I review the specific results for Q4 and the full year, I'd like to provide an update on the priorities I outlined in the last earnings call, specifically gross margin performance, working capital management and capital allocation.

First, gross margin performance. You have seen that our Q4 gross margin sequentially improved and exceeded the high end of our guide by 90 basis points. This was largely due to higher revenue and software mix. We have had constructive discussions with our customers to improve fair value exchange with those improvements appearing in late '26, given the large backlog entering the year. Additionally, we are navigating through particular headwinds from ramping NPI products and rising input costs as supply becomes further constrained due to fast-growing demand. All told, I expect year-over-year gross margin improvements with second half margins being higher than first half margins.

Second, working capital management. We have improved our cash conversion cycle by 34 days sequentially, largely on faster collections and improved inventory days. In fact, our inventory turns improved by 0.4 of a turn. We left the year with 1.4 billion in cash after generating 371 million in cash from operations in Q4 and free cash flow of $326 million. With respect to capital allocation, we completed the first year of our most recent $1 billion stock repurchase authorization, repurchasing approximately $330 million for the year at an average price of $83.34. We invested $140 million in capital expenditures in the business, focused on developing the next generation of leading products and enabling capacity to nearly double our 2025 growth rate.

We also completed the cash purchase of Nubis, supplementing our interconnect portfolio to service the [in] portion of in and around the data center opportunity. We have also reallocated resources that will allow the company to meet the growth challenges ahead with new business processes and technology rationalization.

Finally, let me turn to operating leverage. We will hold to our commitment of flat OpEx in 2026, while investing in new opportunities for our interconnect portfolio. Now let me turn to the specifics of our Q4 and full-year performance. As Gary noted, Q4 revenue reached $1.35 billion, up 20% year-over-year and $70 million above the midpoint of our guide. For the year, annual revenue was up 19% to $4.77 billion, a new record. Q4 was strong across all lines of business. Specifically, our optical business was up 19% year-over-year, driven by strength in RLS, which was up 72% year-over-year. Our routing and switching business grew 49% year-over-year, with the 3,000 and 5,000 series product revenue doubling on a combined basis, with the DCOM opportunity driving much of this growth.

Global Services had a strong quarter, growing 25% year-over-year, driven largely by advisory and enablement and installation and implementation services, which grew 53% and 45% year-over-year, respectively. I'd also like to note that Blue Planet had a very successful year, achieving $34 million of revenue in the quarter, a record $115 million in fiscal '25 and achieving full year profitability. We had three 10% revenue customers in Q4, including two global cloud providers and one Tier one North American service provider. We are exiting the year with about $5 billion of backlog, of which, approximately $3.8 billion is hardware and software with the remaining being software services. This backlog supports a large share of our fiscal '26 revenue expectations, and we see indications of strong demand continuing into '27 and beyond, giving us exceptional visibility and confidence in our outlook and medium- term expectations.

Adjusted gross margin in Q4 was 43.4%, 90 basis points above the midpoint of our guide, driven by higher revenue and software mix. For the year, adjusted gross margin was 42.7%. We continue to mitigate most of the impacts of tariffs as we currently constructed and the net impact of tariffs are immaterial to our bottom line. We continue to monitor the situation and work closely with both our supply chain and customers as necessary. Q4 adjusted operating expense was approximately $409 million and $1.51 billion for the year. Excluding the higher incentive compensation, we achieved in-line OpEx for the quarter and underspent slightly for the year, reflecting our ongoing disciplined approach and operational efficiency. This led to Q4 adjusted operating margin of 13.2%, up 250 basis points sequentially and 320 basis points year-over-year.

Operating margin for the year reached 11.2%, 150 basis points from fiscal '24. We achieved EPS of $0.91, up 69% year-over-year with annual adjusted EPS of $2.64, up a healthy 45%.

Finally, cash from operations was $371 million in the quarter. For the year, free cash flow reached $665 million after $140 million in capital expenditures. Now turning to guidance. Last quarter, our confidence and visibility due to AI-driven dynamics enabled us to atypically provide an early outlook for 2026. As we move into the new fiscal year, for all the reasons Gary and I have discussed, those dynamics and the customer demand environment not only remain robust, they have accelerated. As a result, today I'd like to update that guidance from September as our outlook has improved even from just a few months ago. We now expect revenue in fiscal '26 to be approximately $5.7 billion to $6.1 billion or nearly 24% annual growth at the midpoint versus the 17% growth rate discussed in September.

We continue to expect gross margins for fiscal '26 to be in the range of 43%, plus or minus a point. And as I mentioned earlier, we continue to work to mitigate input cost pressures through supply rebalancing, designing cost out and additional pricing actions over time. We expect the impact of these mitigation efforts will be realized in late fiscal '26. With these dynamics, we expect the margins to improve first half to second half, as cost reductions and pricing actions take hold. We expect adjusted operating expense in fiscal '26 to be flat, at approximately 1.52 billion after accounting for the Nubis operating expenses post-acquisition. With respect to operating margin, we previously advised an acceleration of our longer-term goal of 15% to 16% operating margin from '27 into 2026.

We now expect fiscal '26 operating margins to improve further to 17% plus or minus a point. Our capital expenditures for fiscal '26 are expected to be between $250 million and $275 million. This is higher than our typical capital intensity, in order to invest in supporting expected robust demand in late 2026 and into 2027, as well as incremental costs for 3-nanometer mask sets. In fiscal '26, we expect to repurchase approximately $330 million in shares under our 2024 stock repurchase authorization plan.

Finally, with respect to Q1 guidance, we expect to deliver revenue in the range of $1.35 billion to $1.43 billion, adjusted gross margin between 43% and 44% and adjusted operating expenses of approximately $380 million yielding an operating margin of 15.5% to 16.5%. To conclude, Ciena had a strong 2025, and we are looking to an even stronger 2026. We are thoughtfully allocating our owners' capital to deliver value both to our customers and for our owners. We are singularly focused on executing our strategy and winning in the market. And with that, let me turn it back to Gary.

Gary Smith^ Thank you, Marc. And let me reiterate those comments. We had an incredibly strong quarter in fiscal 2025 which we believe is a seminal one for Ciena, and one that provides a remarkable springboard for continued growth. Our momentum continues to build. Our balance sheet have never been stronger and industry dynamics have never been more positive for Ciena. We are executing well and have high confidence we will continue to do so. We remain very focused on our strategy and continued to deliver the world's best high-speed connectivity, that really underpins today's AI-driven environment.

With that, we will now take questions from the sell-side analysts. Thank you.

Marc Graff^ Yes. Ruben, this is Marc. Thanks for joining. In terms of how much of the new opportunities that Gary talked about are in the guide, they are all in the guide. If you think about the in and around data center, which many of the wins that Gary talked about include, we're seeing nearly a tripling of the percent of revenue from what we saw in 2025 of low single digits to the percent of revenue that we have in 2026 has got low double digits. And so, I think we've captured a lot of that. Obviously, we're continuing to work to satisfy all that demand. But you've seen -- it's all included in there, short answer.

Gary Smith^ Ruben, to the second question around the sort of models around the -- across piece. First of all, I'd say there, obviously, all of these hyperscalers are not sort of homogeneous around their business models, and therefore, their network requirements reflect -- so they are different, notwithstanding, they all need to train. So what we're seeing with the initial hyperscaler is, obviously, just expanding the amount of sites, and that will happen over multiple years. We've had two other hyperscalers now adopt our architecture. So we've got three out of the four hyperscalers have selected us for their scale-across training models. In terms of how -- quantifying how much they are, they are clearly hundreds of millions each but they are different in terms of their scale at this stage.

But really, that's just -- we're just beginning to see the traffic come out of the data center for training around these regional backbones or clusters. We're just beginning to see that. And I don't think that there's going to be a cookie style sort of quantification of the traffic at this point.

Marc Graff^ Yes.

So Simon, let me hit your second question. The three plus 10% customers that we had in Q4, one was AT&T, you'll see that in the K. The other two were not being specific on who they were. But collectively, for Q4, those three coverage just under 44% of Q4's revenue. And then for the full year, it was one cloud provider and one service provider that collectively -- and it was AT&T as a service provider. Collectively, for the year, they covered about 28% of our revenue. Simon Leopold^ Yes.

Marc Graff^ Yes. I'll have to get back to you, Simon. I don't have that specific in front of me.

Are you seeing an acceleration over there? Gary Smith^ I would say that, yes, we're engaged with multiple opportunities with them. Obviously, we're waiting for first GA product, but we have a lot of market engagement, even prior to our acquisition with them, I would say that what we've seen since -- and this is early days, we only did the acquisition last quarter, but I think we've seen sort of accelerated interest now that they're part of a broader -- a broader company.

Scott, do you want to? Scott McFeely^ Yes, just to remind you, they're sort of at a high level, two product families within Nubis portfolio, one is a linear [retimer] that is very effective in terms of extending the life of active copper cable. And we see that as an opportunity that will start in '26. The optical part of the portfolio, the second part of the portfolio, we see more as a '27 and beyond opportunity. But as Gary said, we're getting great feedback from customers on a bunch of different dimensions. First of all, the sort of open ecosystem approach to CPO. Secondly, just the caliber of the team. And then I'd say more an internal reaction. And we -- one of the big things -- one of the big filters as we looked at this company was did we think they are a good cultural fit.

And I'd say 90 days in, we're absolutely thrilled with that.

Marc Graff^ Yes. So there's a couple of things that I would say. The first is one of the big things that we're working on in the first half of the year is ramping our 800-gig pluggables. And so, as we ramp that, you basically get yield economics which will lower the unit costs over time. And we expect that as we go through Q2, Q3 to Q4, we'll see significantly lower cost than we're seeing at the beginning of the year. So that would be the first aspect that I would say. The second aspect is the conversations that we've had with a lot of our customers have yielded good results. So -- once we get through the backlog that we're entering the year with a lot of those new orders will start to see the benefits of those pricing discussions.

So we would exit the year at a higher entry -- higher exit rate then we feel that we'll see in the first half of the year.

Marc Graff^ George, it's Marc. I'll start and then hand it over to Gary and Scott for more color. So a couple of things. You heard me talk about a pretty nice increase in our CapEx year-on-year. And within that, there's about a 50% increase in what we're doing to ensure that we could have more capacity to support what will really be end of year and into 2027 demand. But what we're seeing is really a constraint on the photonics parts, right? And I would say optical parts in general. We've worked really closely with our key suppliers, and I know you know who those are to make sure that we can secure supply. The investments that we've made in 2025 actually yielded a doubling of the growth rate from what we expected a year ago, right?

And so, between that the level of vertical integration that we've got in just what we control, as well as the investments that we're making in 2027, we're trying to support as much of that revenue as we possibly can through '26 and into '27.

Scott McFeely^ And Marc, I'd add to that in terms of the constraints you talked about, in terms of the industry optical component subsegment, if you like. An advantage that we have is a couple of advantages that we have, I think relative to other peers. Number one is, we have a very tight relationship with the cloud providers. We have market share leadership there and a great set of relationships. So even though the demand outstripped what everybody expected, I think we had the earliest view of that in the industry and therefore our conversations with those industry components to started earlier than everybody else. So that gave us that gave us a benefit. I think as we've talked about in the past, we're more vertically integrated than anybody else.

So to some degree, we do have a little bit more control of our own destiny. And part of the reason why in the last 90 days, we've been able to take 2026 out is the activity that we did in '25 that allowed us to double our growth perspective in '25 is carrying over into '26. So we're getting more confident in our ability to deliver to that demand as well.

Gary Smith^ This is Gary. It really varies by specific product areas. I mean, they're all generally -- in the optical infrastructure base, so sort of think Scale Across RLS, they have extended out, but it depends on the product grouping. What we're working on with -- as Marc and Scott said, we're confident -- if you look at the midpoint of our guide which is what, 24% growth for this year, so that's the work as Scott said we did kind of last year to increase capacity and component supply. We're now working on towards the end of '26 and '27 and making sure that we're in a good position from that point of view. So hopefully, by the time we get to the end of year, we get to '27, lead times can come down a little.

But we're seeing increase in demand, including order flows in Q1 being strong as well.

In previous cycles, your margins shot up all the way to even 49%, even over 50%, if we go back a few years and cycles always had a direct impact on gross margin, like you always had a cycle in margin as well. This time because it's coming in pluggables, because it's coming in cloud, your margins are 43%, you're guiding to 43%, 43.5%. And the question is, is there a chance that the margin will also have a -- gross margin will also have a cycle with revenues or what needs to happen for the gross margin to have a similar cycle to revenues? Marc Graff^ Yes. I think how I'd respond, Tal, is there's a couple of headwinds that we're seeing, at least in the near term, and I've talked about the 800 gig.

So we're in an NPI phase of that product. So that's creating some headwinds that we expect to kind of normalize out through the end of the year. The other piece is, I think the customers are starting to see the value in what we're providing, both in space savings and power savings. And we're getting some benefit from that, relative to the value that we're delivering. So I think between those two things, we're seeing that. And my sense is that that's going to be more sustainable than the cyclicality that we've seen in the past because we're really starting to talk about foundation level of benefit that our customers are seeing. And as Gary said, they're realizing that they've underinvested in the network, both on the cloud provider side and the service providers are catching up as well.

And so, I think we're going to see steady improvement to what we've described previously as our aspiration to get back to the mid-40s which at this point, we kind of view as a waypoint, not the end game. So I think we're on a steady track. You'll see sequential improvement, we'll exit the year better than we will perform in the first half of the year, but I'm pretty confident that we're going to see ongoing multiyear gross margin expansion.

Gary Smith^ Yes. I think largely, at this stage, the training Scale Across AI backbones is largely a purview of the large hyperscalers. I think the neoscalers, there's a couple of them that are using the backlog, they're on the back of that for one of a better description. So I think it's -- this is largely right now given the frankly, the scale of it with the hyperscalers. And I don't see that -- I think it's going to take a while for that to bleed through into the neoscalers. In terms of the deployment there, as Marc said, it's going to be this combination of next-generation line systems, RLS, which is also fairly recent into market, and also with the 800- gig plugs as well.

So yes, in the early stages, that's a sort of headwind from a margin point of view, which is why we're kind of guiding as we are. But as that -- as those platforms get more into volume, the yields improve and we get through that NPI phrase on both of those, then we'd have better margins as we exit the year. And of course, you've also got the benefits of just scale and volume as well.

Samik, it's Marc. A couple of things. One is, when we talked about the longer-term guide last quarter, we kind of took those off the table just because in the medium term, we're not very good at calling -- call it, the growth rates on the upside, right? So that 11% to -- 8% to 11%, I think is off the table. We're not really talking to 2027 at this point. But I would say, overall, maybe qualitatively, we feel very strong going into '26. We think a lot of that momentum continues into 2027, and the proof point there is really the increase that we've seen in our CapEx for capacity which is up 50% year-on-year.

Gary Smith^ I think the other thing you could, obviously, extrapolate out. We're not sort of guiding into '27 right now. We're just starting '26. But clearly, the dynamics have changed here. And that's why I said this is a sort of -- '25 was a seminal year for us in this regard. I mean, I think you're seeing multiple Scale Across wins that will -- they are -- by their very nature, they're going to be multiyear. So that gives us confidence in '27 and '28. The other thing I would say is that really a context of our optical WAN-type business in around the data center to it. We're making tremendous amounts of investments and progress on the other dimensions there, we'll give sort of insight and around the data center, which are completely new markets for Ciena.

And the revenues to those are largely -- we're taking some now we're making good progress, largely '27 and '28 plays. And specifically, Coherent inside the data center. That's all additional revenue to us in addition to all the things we've talked about right now. So that gives us confidence in the multi-year dimension to this.

Gary Smith^ Yes. On the DCOM part of that, obviously, that was cocreated specifically with Meta over a period of time. And I think what we're seeing with that is the expansion of the opportunity within the data center piece to that. It saves power and space for them, which is absolutely critical. And we've seen an expansion even in '26. And you're talking hundreds of millions of dollars of this. And as they refresh and build out new data centers, that's become part of their adopted architecture. So I think this, again, is going to be a multiyear opportunity within Meta. I also think about these large hyperscalers really when you think now about the diversity of portfolio that we're dealing with them, they're really markets in their own right, given their size and scale.

We're also, as you said, engaged deeply with two to three other hyperscalers around this kind of architecture. And I would expect to see wins during the course of the year and adoptions for additional cloud players for DCOM. So -- and it gives us an entree point into the data center, together with Nubis, together with the Scale Across, because that is actually -- even the Scale Across is actually installed inside the data center. So you put all those things together, and we're definitely under the tent here. And that's before the Nubis, which will start to deliver product in '26 and before the opportunity with Coherent inside the data center. To your point on the telco piece, I think they've kind of been underinvested in transport, frankly, for the last five years, ever since COVID began.

They didn't want to mess with their networks during COVID. Then you had the supply chain whiplash, and then you had this massive investment they all had to make in 5G, which largely has not yielded the financial terms that they anticipate. Now you're seeing, I think, a multiyear investment back into transport. They are largely underinvested in the networks. And I think whilst that's not at the rapid scale and growth rate of the cloud, I think it's nice, steady mid-digit kind of single-digit growth within the telco space. And I think that's quite sustainable. What is amplifying that though is the AI traffic for things like MOFN. And you saw last year, hundreds of millions of dollars of our telco business was, in fact, MOFN, specifically for hyperscalers.

And you saw it, typically, we've seen it internationally. We're also seeing it now in North America ramp up as well. So you put those two things together, and I think that gives us confidence around that telco, which is half our business currently, having nice, sustained growth rates, albeit lower than the cloud players.

Gregg Lampf^ One more question, please.

Can you give us any kind of perspective there on the product mix? Gary Smith^ Yes. I would say sort of we're laying more tracks at massive scale because it's -- than we would normally see. You've specifically seen that with Scale Across because they're laying the tracks down first. So the much higher proportion of line systems would be the initial take on that. Now that will then move to plugs. We're seeing, obviously, a very large ramp-up in our 800-gig plugs, which we think will be largely adopted amongst most of the hyperscalers. That is a different mix than we've seen traditionally, and more intelligence on the line systems, then also you've seen traditionally, because given the massive scale that they're going to need to put in with multiple fibers across it, then you're going to need to increase the intelligence and the scalability of the line systems.

[Uphold] that with DCOM which, frankly, was a very quickly emerged as a portfolio offering. That was not projected into the '26 plan when we did that in our three-year planning piece. So mix has changed quite a lot around that architecture.

What kind of supply constraints are you seeing for the industry for your cloud customers? Gary Smith^ I think the large -- the relationship with the fiber providers, people at Corning, et cetera, is very tight amongst the cloud players and the service -- particularly the wholesalers, people Lumin who publicly talked about that. So I think there's a lot of commitment to scale capacity. So we're seeing that -- we are seeing that happen. The other thing I would say is it's a real opportunity for us because we have the largest optical support and services organization in the world. And we are increasingly engaged with the deployment of these. In fact, our largest service customer last year was a cloud provider for the first time.

And we're now providing multiple services across the hyperscalers, and we see that as an area of tremendous growth to help, to your point, facilitate the delivery of this infrastructure.

Gregg Lampf^ Thanks, Ryan, and thanks, everyone, for joining us today. We look forward to catching up with folks today and over the next week or so. Happy Holidays and Happy New Year to all.