Ciena (CIEN) Q3 2025 earnings call transcript

Gregg Lampf^ Thank you, [Jamie]. Good morning. Welcome to Ciena's 2025 Fiscal Third Quarter Conference Call. On the call today is Gary Smith, President and CEO, and with us here today for the first time is Marc Graff, who officially joined Ciena as CFO on August 1. Welcome, Marc. We look forward to introducing you to our investment community in person for the coming weeks and months. Scott McFeely, Executive Adviser 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 from the quarter. 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 long-term 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 or included in the investor presentation that we posted earlier today are important part of our 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-K and our 10-Q, which we expect to file with the SEC later today. 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'll allow for as much Q&A as possible today and 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, Greg. And good morning, everyone. I'd like to start by welcoming Marc to Ciena and to today's call. We're excited to have you onboard and look forward to the value and experience you'll bring in leading our finance organization and our global financial strategy moving forward.

Now let me turn to the quarter's results. We had another really strong quarter across the board. Q3 '25 revenue was $1.22 billion, above the top end of our guidance. Importantly, as we focus on driving increased profitability, we delivered quarterly adjusted EPS of $0.67, up 60% sequentially and 91% year-over-year. This really demonstrating our expanded operating leverage from our business. I would characterize demand during the quarter as continuing to be broad-based and durable across both Cloud Provider and Service Provider segments. In fact, we had two 10% customers in the quarter including one global cloud provider and one Tier one service provider, really underscoring our diversified strength and momentum. Further evidence of strong demand was in our Q3 order book, which was, again, considerably above revenue and in fact, sets a new quarterly record for us.

This function increase in orders we've seen in recent quarters really underscores how the Network is now fundamental to the underpinning growth and monetization of AI. At a strategic level, for cloud providers to monetize their substantial AI investments in LLM and GPUs and related data center infrastructure, they need to invest in the network infrastructure that interconnect data centers or they risk stranding their massive investments. Quite simply, AI enablement and adoption is only achieved when data is moved beyond the data center by the network to end customers, whether for training or inferring, these data center investments need to be interconnected by cutting-edge, low-latency, high-speed connectivity solutions. This is a powerful combination that requires the scaling up of system rack density, scaling out between racks within the data center and scaling across with connectivity between data centers.

This will entail a multiyear investment effort on a truly global scale. And this growing focus on high-speed connectivity plays directly to our core strengths and value proposition. Our portfolio, as you know, including WaveLogic Technologies, the RLS platform, the Navigator Domain Controller and our Interconnect Solutions continue to be recognized as the industry standard for AI network infrastructure, solidifying our role as a critical enabler in this transformation. And with an 18- to 24-month lead with WaveLogic 6 and RLS, we clearly have the world's most advanced technology. When coupled with our global customer relationships, this means that Ciena is best positioned to serve these opportunities.

Now let's return to our customer highlights, starting with cloud providers. Cloud providers continue to invest in AI at an unprecedented pace, with many announcing over the past quarter, their intent to increase their expected spend on AI for future quarters and years beyond. Here, I'd like to update you on our progress with two industry-first wins with cloud providers that we signaled last quarter. The first win is for the scale across architecture that I just mentioned. More specifically, this is a dedicated AI infrastructure project related to training and the interconnection of geographically distributed regional GPU clusters. This North American-based project is the industry's first dedicated build for this use case and is comprised of our RLS platform and the WaveLogic 6 Nano 800-gig ZR plug from our Interconnects portfolio.

Initial revenue shipments are underway and we expect this to ramp to hundreds of millions of dollars over the next several quarters. Second of these wins that I'd like to highlight is for a focused application inside the data center for out-of-band network management, a solution we have shorthanded as DCOM. We codeveloped this solution with a hyperscaler, which allows them to streamline the installation and management of its large-scale data center operations, improving scalability and reducing power and space. We also now have significant orders in-house for this application. Overall, we have increasingly strong partnerships with all of the major hyperscalers, driving increased demand for our industry-leading technology in these AI infrastructure builds. In Q3, another major hyperscaler placed its first large order for 400ZR+ pluggables, establishing Ciena as the lead supplier of this technology for this customer.

Consequently, we are on track to meet our expectations to at least double revenue year-over-year for our Interconnect portfolio in 2025. We now also believe that we're likely to be in a position to at least double and more our Interconnect revenue again in FY '26. I want to mention here that there is another sizable emerging group of cloud providers beyond the four to five large well-known hyperscalers. This diverse group of network operators is now generally being referred to as neo-scalers, a term which is inclusive for AI compute specialists, such as GPU-as-a-service provider, cloud and edge service providers and smaller data center and co-location providers. As they build and scale their own infrastructure, neo-scalers are strategically positioned to leverage AI traffic growth distributed compute and automation, creating significant opportunities for Ciena globally over time and adding to the durability of the demand.

In fact, we've already secured multiple new wins with these cutting-edge neo-scalers, and we see this as a rapidly expanding new market for Ciena. Turning now to Service Providers, really on trend with the last few quarters, we continue to see more steady and sustainable investment patterns, both in North America and internationally. And in fact, three of our top five customers in Q3 were service providers. This includes renewed investment in building out their core infrastructure, in part, driven by strong demand from cloud providers for Managed Optical Fiber Networks or MOFN. It also reflects strong enterprise demand pull-through and Service Provider's increasing focus on the role they can play in delivering AI to the edge as more enterprises move workloads to the cloud, and AI-driven applications get adopted over time.

It is clear that both our Cloud and Service Provider customers are focusing their network investments where bandwidth and network scale are critical to support and enable AI traffic growth to drive monetization and adoption. This dynamic is reinforcing the significance of the current and long-term opportunity for both our Systems business and our Interconnects portfolio including over time for inside-the-data- center. To ensure we can take full advantage of those growth opportunities and as part of our regular review of our overall product portfolio, we recently made decisions to align our strategic investments on our Coherent Optical Systems, Interconnects, Coherent Routing and Innovative Solutions like our data center out-of-band management solution, which I mentioned earlier. To that end, we will be redirecting additional R&D investment into these technologies and away from our residential broadband access portfolio, given the larger customer priorities for AI- driven and cloud network investments over the next several years.

To be clear, we will continue to sell and support our existing broadband access products. However, we will be limiting forward investments only to strategic areas such as DCOM. I'll now hand over to Marc for a closer look at our Q3 performance and our business outlook. Marc?

Marc Graff^ Thank you, Gary. And good morning, everyone. Let me start by saying how excited I am to join Gary and the rest of the team to capitalize on the tremendous opportunities ahead for the company. While only a month in, my enthusiasm has only grown as I found my new colleagues, both focused and determined, our technology portfolio world class and the company's cultured center on delighting our customers. Moreover, Ciena and its owners have benefited greatly from Jim Moylan's leadership and guidance as well as a strong financial foundation he's built. As I look to build upon that strong foundation, my concentration will be on driving incremental value creation for our owners. While learning continues, my initial areas of focus will be structurally improving our gross margin performance, establishing world-class working capital management practices and a continuing focus on our capital allocation policies, which will prioritize organic investment in our product and technology roadmap ensuring adequate capital available for inorganic accretion and returning excess free cash flow to our owners.

And I plan to continue advancing the operational efficiencies already underway. Lastly, I will be reviewing our framework and approach to providing long-term financial targets. With that, let me recap our strong fiscal third quarter performance. Revenue of $1.22 billion exceeded the high end of our guidance, up 8% sequentially and nearly 30% year-over-year, with a strong showing in our RLS optical products and Routers and Switches. Adjusted gross margin in Q3 was 41.9%, 90 basis points above our guidance, primarily driven by benefits from sales of previously reserved material and lower net tariff impacts. Adjusted operating expense in Q3 was $380 million, this was higher than expected, driven entirely by incentive compensation associated with strong order performance and our continued strong overall financial performance in fiscal '25.

When adjusted for these impacts, we remain on track to meet our base OpEx spending level for the full year. With regard to profitability measures in Q3, we delivered adjusted operating margin of 10.7%, up 270 basis points year-on-year. Adjusted net income was $96 million and adjusted EPS was $0.67, up 91% year-on-year. In addition, we generated $174 million in cash from operations and a free cash flow margin of 11%. Adjusted EBITDA was $158 million or 13% of revenue. We ended the quarter with approximately $1.4 billion in cash and investments, which included a repurchase of 1 million shares for $81.8 million in the third quarter which brought our year-to- date share repurchases to $245 million. And we expect to repurchase another $85 million in fiscal Q4 to bring the total share repurchase to $330 million for the fiscal year or about a third of our current authorization.

Before I get to guidance, let me provide a brief update on tariffs. Last quarter, we told you that we expect to mitigate most of the quarterly tariff impact and believe that the net effect to our bottom line in future quarters would be immaterial. While still a highly fluid environment, Q3 played out slightly better than expected as we gained more clarity on specific tariffs. We continue to work closely with our supply chain and customers to monitor and respond to any changes in the tariff environment. So as I turn to guide, it is against the backdrop of today's tariff regime and any changes would have a resulting impact on our results. Barring any unforeseen changes, we continue to expect the impacts to be immaterial.

Now on to guidance. For the fiscal fourth quarter, we expect to deliver revenue in a range of $1.24 billion to $1.32 billion. We expect Q4 adjusted gross margins to be between 42% and 43%. We believe fiscal Q2 marked the floor for gross margins, and we expect improving trends over the next several quarters. We expect adjusted operating expense in Q4 to be in the range of $390 million to $400 million, reflecting the impacts of continued strong order flow and overall financial performance on incentive compensation. As Gary noted, subsequent to the close of the quarter, we made the decision to further align our strategic investments toward our Coherent Optical Systems, Interconnects, Coherent Routing and Innovative Solutions like out-of-band Data Center Management Applications or DCOM.

As a result of redirecting R&D investments into these technologies and ceasing further development of our 25-gig PON broadband activities, we expect to record a noncash charge in Q4 and against in-process R&D with a carrying value of approximately $90 million. This will be adjusted from our GAAP reported earnings. Further, as we align our investments and as part of a broader effort to drive operating efficiencies, we are implementing a reduction in head count that impacts approximately 4% to 5% of our workforce inclusive of the broadband investment shift. This will result in a Q4 restructuring expense of approximately $20 million to cover employee severance and related costs, expected to be paid starting in Q4 and additionally adjusted out from our GAAP reported earnings.

Looking further out, we believe demand to be very durable over the midterm horizon in both our Systems and Interconnect portfolios. This is being borne out by our orders and backlog that provide visibility into the second half of 2026. With this strength and the momentum we're seeing in 2025, we have increased confidence to provide a preliminary view of 2026. As we see it today, we expect to deliver approximately 17% year-on-year growth in fiscal 2026 similar to what we're currently projecting in fiscal '25, achieving the high end of our three-year revenue CAGR target one year early. We expect that gross margins will continue to improve in fiscal 2026 with an initial estimate of 43%, plus or minus one point.

And we expect to continue investment in our product and technology roadmap funded by a combination of portfolio decisions and operational efficiencies. The net result of which will enable fiscal 2026 OpEx and to be flat to fiscal 2025 at approximately $1.5 billion. Taken together, we now believe that we will accelerate our longer-term goal of 15% to 16% operating margin by one year from 2027 to 2026 driven by increased operating leverage and improving gross margins. Given that, and in combination with the continued rapid growth we are seeing in the market, we will not be providing new three-year targets.

With that, let me turn it back to Gary for some closing comments before we take your questions.

Gary Smith^ Marc. In summary, we are seeing strong broad-based growth across our portfolio, driven by accelerated demand from AI workloads, related cloud traffic and steadily improving service provider investments. To be clear, we believe we're in the midst of a positive secular change in the scope of our opportunity, as well as our growth trajectory, given that the network has never been more critical, particularly as traffic flows out of the data centers. And as you know for the last several years, we've been investing in technologies and products that address this demand. That demand is here now and it's only just the beginning. Globally, more than $7 trillion is projected to be spent through 2030 to accelerate investments in data centers, GPU clusters, the power grid and AI model development, and increasingly, networking will have a greater wallet share of this spend.

We believe we are at the early stage of a multiyear, highly durable network investment era. It's now all about the network and Ciena's high-speed connectivity is front and center in this critical AI scaling, as it basically moves from electrons to photons, it's all about optical.

With that, we'll now take questions from the sell-side analysts.

Gary Smith^ Thanks, George. Listen, I think as you know we've invested very strongly in the high-speed connectivity, and we clearly have an 18 months to 24 month lead on that. So I think that's highly valued by the cloud providers and hyperscalers and service providers as well. So I think we have a very strong competitive advantage there. And I think the structure of the industry has improved considerably over the last few years. And that does play through to our expectation. It's a confluence of elements around our improving gross margin. I'll let Marc talk to some of the other elements there. But basically, that's part of the economics that we believe we can get over time into the mid-40s gross margin.

As Marc said, we called really Q2 was the bottom for that, and we expect to see steadily improving gross margins quarterly over time.

Marc Graff^ Yes, George, this is Marc. What I would add to what Gary said, is we're really trying to take a pretty structured approach to how we think about gross margins, and we think about those really in three buckets. The first bucket is how do we design products that have a long life with continuing reducing cost structures. The second is optimizing how we work with our supply chain and our suppliers to really make sure that we're realizing and accelerating those unit costs on those items that we procure. And then the third piece is really making sure that we're having the conversations, which we are real time with some of our customers to ensure a fair value exchange.

As we think about the current construct, we're really excited about expanding our footprint, and given the long life of these products, we think that sets us up really well for gross margin expansion as they continue to invest in their networks.

Gary Smith^ I think the expression value exchange is a much nicer expression, George.

Yes.

Gary Smith^ Samik, listen, I think it's a net incremental opportunity for us that's going to scale over time. We're encouraged by what we're seeing at the early stages of that. A lot of these neoscalers are identifying fairly early on that they want to invest in the network. They understand the issues around inter-connectivity and how they're going to enable their business. So you're seeing a lot of new builds beginning with these neoscalers and we've won a number of them, and we have a number in our pipeline. So I think as we go through '26 we've included that, obviously, in our guide, what our expectations are for these. But I think they will increasingly, over the next few years, play a more significant role in our overall growth.

Gary Smith^ I think increasingly, obviously, it's driven by AI workloads. But I would also say that we expect service providers to play an increasing role in that. Both delivering hybrid networks to the cloud players, and also in these managed optical fiber networks. We're seeing multiple opportunities both in North America and internationally in MOFN. So roughly, right now directly, hyperscalers in their various forms including neoscalers, is roughly about 50% of our business. Our kind of view is I expect that to be similar actually in '26. But a large part of that service provider workload will be MOFN and cloud and AI Edge related. So I think it's going to be a confluence of platforms that deliver those workloads.

Marc Graff^ Yes. I would just -- Samik, it's Marc. I would just add that we've got reasonably good visibility into that, given the huge backlog that we have today and what we expect to leave Q4 with. So I would agree with what Gary is saying.

Marc Graff^ Yet, it's Marc. Let me take that and then I'll let Scott and Gary add some color. In terms of the gross margin uptick that we're looking at, as we scale those parts, we're expecting, obviously, those unit costs to come down. And that really got the way we expected in Q3. As we look forward to the 42% to 43% guide that we gave and then the 43% plus or minus guide that we gave for '26. We expect those trends to continue. And so, we should be seeing some tailwinds from improving ramps of those products as they get costs reduced just through scale. The second part of your question around the goodness that we saw in Q3 from tariffs.

We told you last quarter that we expected to be about $10 million roughly in tariff costs per quarter. As we saw the administration kind of work out some of the trade deals with various nations that impact us. Like India, Vietnam, et cetera, we're able to tighten up some of those expectations. So we saw a little bit of good news, maybe it's 20 or 30 basis points of the benefit that we got out of the $90 million that I cited. And that's really -- yes, that's really what was driving it was just getting tighter on a uncertain environment.

Scott McFeely^ And Meta, just to follow up on your margin question. You referred to headwinds that we referenced in previous calls around some of our new product introduction, whether it be pluggables or 6ZR or RLS. And Marc gave you the dimensions of what the margin changes were quarter-over-quarter. But I would say we did that margin improvement, despite the fact that we were shipping a lot more are less and a lot more RLS and a lot more plugs and to give you a sample of that quarter- over-quarter on their plug business that ports that we shipped were 20% sequentially. So we were able to deliver margin improvement even in the face of that and RLS had another bumper quarter.

And that's all coming from scale in general matters, the natural life cycle of these new introduction products and our continuous focus on cost reduction over time. And then as we look forward into '26, we will be able to take advantage of a technology transition into our WaveLogic 6 family as well which will yield better margins for us.

You're absolutely correct. The scale across network, as you say is sort of connecting GPUs at distance beyond a single data center. The actual specific example that we referenced in terms of the first customer movement on that, at a high level, you would have thought that may have been an opportunity for Coherent Lite, but what we actually see is more and more often, these cloud providers are needing to bring the network to the power and the power and where they can get the power is dictating the distances and the performance requirements in order for you to be a leader to service that demand, you need to have a breadth of capabilities, Coherent Lite be an important one of those on the menu.

But this particular example, we had to actually use our Performance Optics. So it's 800-gig six Nano because of the distance they were pushing. But for sure, Coherent Lite still believe in the opportunity is not even a question of if it's a question in. We're -- the view we put out in the past was that we would expect to see revenue start to flow on those technologies in '27. I don't that our perspective has changed from that.

Gary Smith^ I think, Ruben, it's more the latter approach. I think if you think about the growth opportunities that we've got, across the board on our AI workloads, we really want to prioritize those. So we're going to continue to support our broadband initiatives there. We are just really not going to put the longer-term roadmap for that in place. which we were embedding in. And I think given our priorities, it makes absolute sense to -- but it front and center behind the optical portfolio.

Scott McFeely^ Yes. And Ruben, specifically, when we say defocus or prioritize the future investment. What we're talking about here is the higher capacity PON technology is 25-gig and 100-gig PON. PON is still an important technology for us in terms of offering an access choice on our Routing and Switching portfolio and a key contributor to optics inside the data center with our DCOM solution set.

Or is the strength this telemetry opportunity that Gary mentioned earlier in the call? Gary Smith^ Simon, let me take the first part of that, and Scott will take the second part of the question. I think what's different around this data DCI connectivity we're seeing, it's shorter distance, and it's a dedicated training network, if you were to put it simply. So it's super high speed, low latency. That's why it's using our RLS and the 800-gig plugs at scale. This is just one region of this particular hyperscaler. We believe that there are other applications that will -- this same application will roll out both at this hyperscale and at others. It's the first one of its kind that we're aware of, and we're beginning to roll that out now but it's dedicated for training across multiple GPU clusters.

Scott McFeely^ And on the writings question, Simon, I think there's two dynamics here. One, for sure, it has mirrored sort of the comeback in terms of service provider spending. As you know it was largely exposed to the service provider space in terms of the opportunity set. So that that's one key driver. The second one, though, is we are seeing an emergence of Coherent Routing opportunities. This is playing to the strength of our optics inside our routers and you start to see that in service provider aggregation networks. Then the third one you mentioned it, which is the opportunity to take that Routing portfolio and the technologies that we have and create this optics inside the data center for their communication network.

So the DCOM, as we called it, opportunity. So those are the three drivers, and they're all contributing.

Scott McFeely^ Yes. Let me tackle the first one on the interconnect business. Just a reminder for folks on the call when we talk about interconnect, it's really the categorization of selling our speed optical technologies, either as plugs or subcomponents independent of our System business, so that's what embodies the interconnect business. As I said earlier, we had our best quarter ever in terms of plug shipments. They're up 20% sequentially and up something like 140% year-on-year, which puts us well on track to beat or exceed our stated goal of doubling that piece of the business from '24 to '25. We are shipping revenue against our 6-N, which is the next technology platform, both in terms of complete plugs, but also subcomponents of that as DSPs and our electro-optics sold as component unbundled from our plugs, and that's just starting.

I would say the number of customers here measures in the 10s, multiple 10s. The confidence that we have to say we are capable of doubling it year-over-year actually is in our order book. We have a fantastic backlog in order book here. We got great visibility into next year.

Gary Smith^ The second part of your question around the application I talked about with the dedicated training network. It is multiple hundreds of millions. It is a considerable amount of RLS, which is pretty much the de facto standard now amongst all of the hyperscalers, our line system. We're laying a massive amount of tracks around the globe. This is another example of it. We have the orders for this first region in-house. We are beginning to deliver in Q4, recognized some revenues in Q4 with fairly small. But it will ramp up in Q1 and Q2 of next year. That is really just one region, and it's the first time that we've seen this application -- and this obviously is a new application for connecting data centers.

It really represents, we think over time a pretty expansive TAM opportunity for us. And obviously, we're super well placed for both our line system, our modem technology and, in fact, our domain manager, they're all industry-leading, and you put those things together, and it's -- it really is sort of de facto standard for this kind of high speed, low latency application.

Marc Graff^ This is Marc. You had a question about the impact on the margin. The 43% that plus or minus that we gave you for 2026 is inclusive of what we have in place today as we continue to ramp that, one of the benefits that we see of these large orders is it gets us to scale faster. And so, that will be a tailwind for us as we continue to ramp those solutions.

Scott McFeely^ Thanks, Tim. I assume you're talking about WaveLogic 6 Extreme. If that's not right, you can jump in and correct me. Great success with 6-E in early days. We added 11 new customers last quarter up to a total of 60 customers.

It doesn't really seem to be impacting you, but I wonder if you could comment on the overall component or supply-constrained situation? Gary Smith^ Yes, Tim, I would say that we've invested significantly across the supply chain ecosystem, and we continue to do so. That's really what has enabled us to grow year by quarter comparison, that's a result of that 17% midpoint of the guide for the year. So you can see we've scaled that up. We have made considerable investments to continue that scaling up in 2026, given the opportunity and the sort of, as Marc indicated, we expect similar kind of growth rates next year. Obviously, the scale of the business, we're at, you can imagine that requires considerable investment, which we've already made.

So we believe that whilst we're constrained, our revenues would be higher if we've had bigger capacity, but we are ramping that up in alignment with what we see the market opportunity to be. So we still have challenges in certain components, as you ramp up this kind of scale, but we're working our way through them. And we believe that the things that we've put in place across our global supply chain will put us in a very good position to, over time potentially, as we get through '26, reduce some of our lead times as that capacity comes online.

Gary Smith^ No. Amit, I think there's a big change going on. And it's not just, obviously, you're layering on Service Provider is continuing to have very good sustainable growth. If you think about the Service Provider market, it's really under-invested in core optical infrastructure over the last five years, A, because of all the COVID supply chain, absorption and all the rest of it. And B, the preoccupation with 5G. They're now I think well over that. And we're seeing, I think, a good runway for just service provider growth. It's at a lower growth rate. It's in the mid-single digits, but it's super helpful, it's half of our business. So we have pretty good line of sight to that.

And I would also say that three out of the top five customers this quarter were service providers. And I also think the other dynamic around this shift of workloads, as you start getting these AI workloads closer to the edge, you're really talking service providers. And that in addition to the MOFN opportunities, which we're seeing both in North America and globally continue to expand. I think service providers are going to play a critical role in this whole AI workload delivery. So I would expect, even though we're going to have outsized relative to service provider growth in cloud, I think the mix is going to be fairly similar in terms of direct next year. So they're both growing.

Hyperscale is, obviously, at a higher rate, given the global sale of the investment necessary for the data center interconnect.

And if that's the case then, should we think about Ciena having perhaps better market share which you've historically had going forward in this cloud opportunity? Gary Smith^ I think that's a very fair observation. If you look at RLS. RLS was actually also a collaboration with hyperscalers in terms of the design to it. We're actually leaning into them across a number of technologies where we're co-creating the technology and the evolution of both the line system and the modem technology as well. So you look at the kind of growth rates we're having here. And I think over time one can now only conclude that we're going to continue to expand our market share in this space.

Gary Smith^ David, on the DCOM side, this is one customer that we've co-created with. We think – well, we know that this application can be generic to data center technology given that it plays nicely to reducing power and space. which is critical. I would say just with this one customer, it is hundreds of millions. And we've got the first initial orders for that, and we will begin deployments shortly, but that's continuing to ramp. So it is hundreds of millions and that's just one customer.

Marc Graff^ Yes. So again, as you think about the impact to the margins, David, as that thing ramps, again we get the scale benefit and that scale benefit kind of ripples through the entirety of the P&L. So like Gary said, we're starting to ramp that now. This particular customer is one of our largest customers for this quarter. This particular workload, DCOM is starting to ramp within that customer's envelope. And so we expect to get some tailwinds from that from a margin perspective as it ramps through Q4 and then, obviously, to hundreds of millions, as Gary talked about in 2026. David Vogt^ Then overall fiscal ‘26 kind of color, if you can share?

Marc Graff^ Yes. So, it's really around the backlog that we're seeing this year. right? So as we talked about in the previous question, if we had more supply. We're able to get more revenue. As that supply kind of goes into 2026, those orders are going to get fulfilled in '26, and I'll recognize that revenue then. So as we talked about the visibility into the second half of -- we've got fairly high confidence that the guide that we gave you is going to be achievable. As Gary mentioned earlier, we're also seeing some pretty healthy growth as witnessed again by orders from our service providers. So the combination of those two things from both our cloud side of the customer house and the service provider customer house, is giving us quite a bit of confidence as we go into 2026, which is why we decided to give you the 2026 guidance a quarter earlier than we normally would.

I'm not sure we'll do that again. But given the confidence that we had and the momentum that we're leaving this year with, we felt pretty comfortable giving you guys a sneak peek on what 2026 would look like.

Gary Smith^ The other point I would make is as you look out longer term here to the durability to it, first of all, you're looking at massive scale of investment that's going to be necessary. So I think we will take comfort from that. Secondly, you've got a lot of these new applications, some of which we've talked about today. But we also -- none of this sort of guide that we've given you includes really inside the data center coherent adoption, which we think that intersection as you move from electrons to photons as you require less latency and higher speed is going to happen. It's just a matter of when. And so, that's all in front of us.

So not just the '26, but beyond, I think we have a very good high confidence in the dynamics that we're seeing. There's a secular shift in our opportunity.

Marc Graff^ They are. So sorry, maybe I wasn't super clear. So one of the things that we're looking at is pretty significant increase in operating leverage as we go from one 17% growth year to the next 17% growth here, while we're keeping OpEx flat. So what I mentioned in the prepared remarks was we are pulling in the 2027 15% to 16% op margin goal into 2026. All right? So I want to be super clear. It's a one-year acceleration. Our expectation is we hit 15% to 16% in fiscal 2026. Does that help, Karl?

Obviously, it's a little early to talk about '27. It's a little early to talk about 26%, but we have good visibility into it. So we're giving those. Listen, I think this is a very durable multiyear secular shift in the demand environment. And when you think about it, so much investment has gone into AI in its various forms. The GPU accelerators, the LLM, data centers and all infrastructure about that. Now it's time for the network because without the network, it's got to come out of the data center to enable all of this stuff to happen. Monetization, training, inference, all of those things need the network. There's a greater prioritization right now in terms of driving the growth of that network -- so you don't strand these assets and you really enable the whole AI infrastructure.

We're at the very early innings of that.

Gregg Lampf^ Thanks, [Jamie]. We have time for one more question.

We are in terms of capability, I believe, the most vertically integrated supplier in the industry. That's not to say that we depend only 100% on our own components because we -- from a strategic perspective, we use our own component and we also use third-party component providers to provide us flexibility. In a constrained environment like we are now we're using all those levers.

Gary Smith^ I actually think we're seeing both, frankly. Because of our leading technology and the investments we've made, the very focused technology spend on line system, modem, domain management. We're 18 to 24 months ahead of everybody else on that. Obviously, consolidation is a helpful dynamic in the industry as well for sure. But you're seeing, I think at a very similar level, you're seeing service provider return to spending on infrastructure after a sort of underinvestment in for five years. So that's a nice dynamic to have and then you're leaning in on all the AI workloads. It's all about the network and optical as it moves from electrons to photons, both in the DCI market and eventually in the data center kind of plays to our strength, and that's why we think this is a very durable multiyear fantastic demand dynamic, and we are incredibly well positioned to it.

The investments we've made in the last few years really intersect all of the key elements around this high-speed connectivity that's going to be required.

Gregg Lampf^ Thank you, everyone. We look forward to catching up with you later today and over the coming weeks. Thank you.