We'll get going here. My name is Alex Henderson. I'm the networking and security analyst at Needham. been following Ciena for geez, 25 years plus, something in that range. and it's a pleasure to have them here at the Needham Growth Conference. Today, we're going to do a fireside chat with David Rothenstein, who's the Senior VP of Strategy at Ciena, and therefore uniquely positioned to talk about the broader trends, and demand and supply dynamics and things of that sort. Thank you for joining us.
Thanks, Alex. Thanks for having me be here.
Let's start off, for people who don't know the company, you had a fabulous October quarter. You had a horrible July quarter. You know, can you talk about the last two or three years trajectory into that and where we are today looking in the rearview mirror just to make sure everybody's on the same page?
Sure. Thank you for the blunt honesty, Alex. Good afternoon, everyone. Yeah, I mean, it's been a choppy several years for Ciena obviously, and for everyone. Particularly going into the pandemic and then, the past 12 to 18 months of significant supply chain constraints have brought about a lot of different changes in our industry and in our business. The way to kind of think about it over the past three years, after a couple years of hyper growth in 2018 and 2019, the past three years have been largely flat in terms of top line growth for the company.
That was, you know, for a lot of different reasons, but particularly with COVID in the first few years, you had network operators really constraining their spending across the board. We've been around $3.6 billion the past three years. In 2022 in particular, we started seeing some significant impact from supply chain challenges, initially from semis broadly. Then in Q3 and Q4, Alex, what you're referring to is some specific challenges with availability of certain integrated circuits, particularly Power Management ICs, that really gated the manufacture and shipment of modems for us.
What you've seen is, even though the vast majority of our supply chain is actually performing to what they, you know, have told us in terms of volume and in terms of timing, the volatility of a small number of component suppliers for these integrated circuits can have significantly unfavorable effects, like it did in Q3 or, consequently, correspondingly, favorable effects in Q4. What we are seeing kind of as we're into our Q1 now, we're on November to October fiscal year, is continued improvement from a supply chain standpoint. We are seeing more and more suppliers delivering what they said they're going to do, but still some degree of volatility amongst those.
We have been able to start, seeing the benefits of a lot of our mitigation efforts on the supply chain side in terms of qualifying alternative sources of supply, having some product engineering redesign efforts come through, as well as improvements in production capacity. All those mitigation efforts are bearing fruit and led us to, you know, result in Q4 significantly above our guidance.
Looking at the mechanics of that, a single decommit late in the July quarter, uncertainty around the October quarter, your guide there, by almost by definition, has to be fairly conservative because you're not sure what it's gonna look like. Then better than expected results. Is that a function of them having decommitted and coming back in and, you know, if I average those, that's the rate of supply that you're getting? Or is it that they're, in fact, seeing better availability and more consistency, so you'll actually see sequential gradual improvement, as we go forward quarter to quarter to quarter?
I'd love to say we had complete line of sight in terms of all those different dynamics and in terms of how some of these suppliers do their own allocations, particularly among industries. You know, I do think that certainly in the middle part of last year, there was some allocation heavily toward other industries. With some of the, you know, higher inflation and recessionary impacts, could that have an impact in terms of allocation going forward? You know, I think directionally, perhaps, but I wouldn't take that to the bank. You know, I think overall, you know, we can control what we can control, which is, as I said, really trying to continue to drive the mitigation efforts and offset whatever ongoing volatility we'll continue to, you know, to instantiate.
What I do think is that overall more broadly is from a supply chain standpoint, I think we are going to see continued improvement, but it's going to take time. There's not going to be a switch that gets flipped in one particular quarter or month that says all of a sudden supply chain issues are behind us, and we can now just focus solely on demand. I think it's going to be a multiple quarter effort.
Well, probably a multiple year effort at that.
At potentially multiple years. That's right.
Yeah. Looking at, the backlog, just to remind us, you entered, you know, a couple of years ago, your normal backlog is what, four-eight weeks?
The normal product delivery times are, you know, have been before, you know, before all the supply chain issues were four to eight weeks for the majority of our portfolio.
What is it currently?
You know, the average of the portfolio is around eight to nine months right now.
And.
With variation.
As supply improves, the duration of the orders should come in?
Yes, as supply improves, you know, I think a couple of things are going to happen. One, I think you're gonna start seeing our extraordinarily high backlog start coming down, which, by the way, is a very good, healthy thing. It's, you know, we entered this fiscal year at $4.2 billion, which is a historical high for us, not just in terms of absolute numbers, but in terms of relative, in terms of percentage of our demand plan for the year. It is almost the entirety of the demand plan for the year, which is highly unusual, and it's a function of, you know, the longer lead times and frankly, the forward ordering by several of our larger customers and the hope and expectation is that as we start to service that demand, backlog will come down.
Order growth, I wouldn't say it's gonna slow down or decelerate. I think it would normalize to where it, you know, can and should be going forward.
When portfolio managers hear the word massive backlog, biggest in the company's history, there's almost a Pavlovian response to that, which is, "Oh my God, I'm selling the stock.
Yeah.
That's probably been borne out of history. A lot of semiconductors and other component companies, have seen significant corrections in their backlogs a s double and triple ordering have evaporated. That's not the case with Ciena. Historically, you've almost never seen backlog be canceled. Can you talk about the stability and predictability of actually shipping what you have in backlog? Because it's a critical piece of the puzzle.
Yeah, it's a really good point, I would caution anybody in terms of reading through, you know, the way semis are ordered and procured, versus system solution sales, which is what Ciena does. Now, the reality is, our products, while the majority are commercial off-the-shelf, they are highly bespoke. You know, we have over 11,000 SKUs in our product portfolio. In the vast majority of cases, the solutions are specifically configured for a particular customer's network architecture or network application. In many cases, it can be a multi-quarter effort on the part of the network operator to actually run an RFP or an RFQ and make a selection process and then do the network design as part of that.
You take that on top of the fact that two of these are incredibly expensive system solutions. These are not jellybean commodity parts that can be easily replaced. Three, you know, given the fact that during the pandemic, many of the operators did run their networks particularly hot, you take all three things together with the fact that the vast majority of our order backlog is, frankly, contractually non-cancellable. Even if it weren't, I'm not sure it would change the dynamic. You take all those things together and you say, "You know what? There really isn't where I think you're going, any material risk to degradation of the existing backlog.
Going back to the point of availability causing duration stretching that caused your backlog to go from $1 billion to $4.2 billion over the last two or three years, the reverse is obviously gonna happen at some point during the next 18-24 months. As that happens, how do we think about what happens when you get upside availability on components, then the duration by definition narrows, therefore the time they have to order ahead diminishes, and therefore the orders decline. There's kind of a multivariable equation there that's a little bit difficult to solve. If we assume, say, 5% or 10% upside due to the revenues, just hypothetically.
Y ou'd still have orders that are reasonably healthy in a typical year that would allow you to bring that backlog down, maybe bring it down a couple, two, three months, but you'd still exit 2023 with a backlog significantly above normal. Is that mechanically reasonable?
I think the thesis is absolutely correct, Alex, and frankly, your view in terms of having to thread the needle in terms of some of these forward expectations and projections is correct, right? I mean, I would keep pointing out, though, that all of this is frankly healthy, right? The backlog is inflated right now for the reasons that we know. Our inventory is inflated right now for the reasons that we know in terms of building up materials, waiting for the golden screw to turn it into finished goods. All of that is and should normalize over the course of time, and that's a good thing. Where I think you're going is, you know, where is it going to end up? Anybody's guess. I think certainly for fiscal 2023, I don't see a return to status quo ante.
I don't see backlog returning to where it was before supply chain challenges. I don't see inventory returning back to where it was before supply chain challenges. I think it'll start normalizing. You layer that against the fact that we, like, frankly, any global multinational who's worth their salt, are rethinking their supply chain operations going forward. Instead of this 25-year journey toward just-in-time supply chains focused on operational efficiency and cost, we're seeing and looking at how we balance that with a greater degree of resiliency and redundancy in our supply chain, which will likely need some degree of higher inventory to take account of some of the risks that exist.
Embedded in all of this is a need to get parts that are often almost impossible to get, and what parts became available are often at incredibly inflated prices. How much cost are you absorbing currently as a result of the supply chain and logistics challenges that once it normalizes, will come back out of the margin structure?
What we've said about margins is, you know, we've said it for quite some time, actually. Our long-term target margin is in the mid-forties. We did see at the beginning of the pandemic, you know, several quarters of very high margins in the high forties as a result purely of product mix. I mean, the reality, Alex, any question about margin, the answer is product mix one way or the other. We saw virtually no operators undertaking new builds, and as a result, a huge proportion of the sales were capacity adds which carry higher margins. That has now, you know, regressed a bit.
Specifically, with the supply chain challenges that you're talking about, what we've seen is what we said, generally with the semi-shortage is about 400 basis points of impact to the margin line. That's a function of a couple of different things. One, it's instead of the typical cost reductions we would get out of the supply chain, we've actually seen cost increases from some of our component suppliers. As well as the fact that we've seen the need to pay, to your point, higher premia to brokers on the open market to procure alternative components, as well as, you know, for a time, higher shipping and logistics costs. It just cost more to get product anywhere around the world, given the shipping lane challenges that we had over the past year or so.
All of those things, as I said, about 400 basis points. On top of that, there was maybe an additional 100 or 200 basis points of impact from the specific integrated circuit challenges that we've had. Now, we've said for 2023 is we've got it to a margin of around 42%-44%, which is roughly flat year-over-year. We do expect, and we are starting to see some of those supply chain challenges, as I said, improve, which is resulting in slightly lower costs, but they're still elevated. What we've said is around 200-300 basis points this year of margin impact because of these ongoing dynamics.
Theoretically, that could fall out over, say, 2024 and into 2025 timeframe.
We think it will. As we've said, we expect in that timeframe to return for our target mid-40s margin.
Yeah. In addition to that, it sounds like there was a pent-up demand for footprint build, which obviously the chassis are much lower margin than the blades and transceivers. To that extent, are we seeing a mix that stays that way over the next, you know, year? Or is it going to start to normalize on the mix as well as we exit 2023?
You know, I think there's a lot of different dynamics at play. You know, we had, both in Q4, what we're seeing at least the beginning of Q1, with the increased availability of some of these integrated circuits, our ability to manufacture and ship more modems than we had certainly in the middle part of last year. We're also starting to see, and it's one of the reasons why we guided to margin the way we did, a shift in product mix, where we're starting to see more and more, line systems and early builds and new deployments coming online, which is great because now we're starting to see some of these new wins we've had over the past few years come to fruition, which is a good thing. It's a healthy thing.
The mix shift, again, trying to predict it quarter-over-quarter is super difficult exercise. But we do see that dynamic starting to normalize a better, a bit better.
Based off of what you're saying, it sounds like if supply, if the economy decelerates, you know, most, I think, portfolio managers would argue that the S&P industrials is, you know, gonna earn $235-$240 in forecasts, but actually the estimates are all too high, and they're gonna have to come in to $200-$210. That's the mindset that's out there. The Street's bracing for estimate cuts. If the economy decelerates, you would think you would get more parts, maybe even at a better price, resulting in potential upside to your revenues, some upside to your gross margins, and you're leveraging your OpEx, so you could theoretically beat and raise because the economy slowed. Logically speaking, that's a reasonable premise, yes?
It's not a bad syllogism. I, you know, I think a couple of things. Yes, you know, we have postulated that with, you know, potentially the higher inflation rates and recessionary impacts certain industries, like for example, consumer electronics, could that be an opportunity for some of the key integrated circuits to be allocated to other industries, including our own? I will tell you, we're not privy to exactly how the component suppliers allocate among industries, so I can't predict exactly how that has happened and will happen. You know, there's certainly any more allocation that we would get would certainly go a long way toward accelerating a return to some degree of normalization on supply chain. The flip side, of course, is trying to predict what impact current macro, you know, dynamics will have on any particular customers. CapEx, you know, decision-making is a different story.
Ironically, an industry that's known to be, you know, a cyclical industry actually looks like it's because of the supply chain problems over the last, two years, may actually be countercyclical this year.
Yes. There's some irony there.
Which is kind of ironic in there, right?
Yeah.
Let's shift over to the demand side of the equation. That's obviously an area that people are concerned about. Why don't we start off with the cloud side of the business? There's been a fair amount of discussion about potentially some of the cloud companies having overstocked inventory. There's been some comments from some of the cloud companies that they're going to slow data center builds and maybe even cancel a few of them. Have you seen any of that? Do you think you will see any of that, you know, not just in 2023, but as we go through 2023 and get the orders for 2024?
Well, like you, Alex, we're certainly watching very closely in terms of what our customers are doing and saying, and there's been a fair amount of media attention as of late in terms of things like mass layoffs and travel freezes and hiring freezes and talk about different spending changes. Honestly, at this point, we are not seeing any of that have any visible impact to us, not just now, but going forward. You know, given the size of our backlog, this really isn't an issue for 2023 for us.
Certainly, in terms of what our sales account teams are saying, who are closest to our customers, at the end of the day, I think what's really important to remember, Alex, with any customer, it's not overall what they're doing or how much they're spending, it's where they're spending their money on.
Exactly.
Where their investment efforts and intensities are going to. As of right now, we see with the cloud providers, and they're not a homogeneous market segment, I know that, but I'll talk about them as one for now. We are not seeing any change in their demand profile at all, because the reality is, the amount of cloud applications and cloud services they're gonna need to host on their platforms continues to proliferate. They're gonna continue to have to refresh their data centers. Yeah, they may slow or cancel a data center build in the future, but that's a leading indicator. It doesn't change their demand needs right now. What we're seeing is they are continuing to invest in their cloud infrastructures. Their cloud revenues are up. I mean, some of them, their ad revenues are down, but cloud is up, so.
Yeah. What was AWS's growth last quarter? Was it 37%?
It was.
Oh, by the way, their backlog went from $100 billion to $105 billion.
Aren't too many companies who wouldn't take that.
I think that's the largest backlog in the history of anything. I mean, I don't know if there's ever been a larger backlog. What about on the service provider side? Obviously, you know, people can make their decision on cloud. I think it stays reasonably healthy and continues to grow, at least at the rate you guys are forecasting growth at. Service providers spending is more constrained to start with, and their business model is more constrained. Have you seen any change in their behavior, and how should we think about what they're doing? Verizon in particular was mentioned as a company that's talked about, you know, cutting back on CapEx a little bit as we go forward. How are we seeing them there?
I think, you know, I did deliberately talk about cloud provider homogeneously. Service provider can't be spoken about in that way because tier ones in North America are fundamentally different from Tiers in 2s and 3s versus international service providers. I think it's hard to generalize, Alex, in that regard, but if you're insisting that I do so, I think by and large it's the same answer. We are not seeing really any significant shift as of right now in CapEx spend. Could there be a push out of a route or a build here or there? Absolutely. Could there be a redirect of a route to a competitor who's offering a slightly quicker delivery lead time? Sure.
By and large, the discussions that we're having, and we rely heavily upon our sales force, who are as close to our customers as anybody, and in fact in many cases know our customers' networks better than they do, are continuing to say and believe that the demand is there. The demand is robust. Just like the cloud providers have demands upon them, the service providers are gonna have to continue to invest to figure out a way to modernize and monetize their network assets at the edge to automate connectivity. That is not going to change. You could have, you know, here or there. You mentioned Verizon in particular. It's a good example.
Verizon, you know, to quite a lot of fanfare, took their CapEx down year-over-year. It was actually inflated in 2022 because of spend on wireless spectrum. In fact, their spend on cloud infrastructure is not going to change. Again, it's another example of where they're spending is what's really important to think about when you're thinking about Ciena.
One of the areas that people don't spend a lot of time thinking about is the subsea market, which are three, four, five-year projects. You guys have had a lot of success in subsea. It's what percentage of your business now? What, 15%, 20% of your business?
It's a bit lower than that. Submarine standalone, and again, we only play in the SLTE, the submarine line terminating equipment is about high single digits.
High single digits.
8%-9%. You know, it is a really important proxy as well for the cloud providers because, you know, we talk about cloud providers, in terms of direct data center interconnect, but cloud providers also come through in terms of carrier managed services, where the carrier owns the fiber and provides a managed service for the cloud provider, which shows up as carrier revenue and subsea, right? Where in many cases the cloud providers are now being the primary dictators of where that submarine, you know, cables are going and where they're being upgraded. We feel very good about subsea. We see a significant number of new build opportunities.
Yeah.
Out there, you know, around Africa in particular, ones that are interesting for us. We're very excited about the subsea space.
The access side of the business is an area that you've spent a lot of time building capabilities in and expanding into somewhat of a new TAM, and that's growing much faster than the company as a whole. Can you give us some sense of the size and growth rates of demand in that side, and what have you done to create that opportunity?
Yeah, I mean, it's a good question. I mean, I think the way I think about routing and switching more specifically, you talked about access. It's a component of it and we refer to it more colloquially as Next-Generation Metro and Edge, which is, you know, covering everything from fiber broadband access to cloud connectivity, software-defined edge, and of course convergence of IP and optical. All of that we see as far and away the fastest growing TAM expansion opportunity for Ciena over the next several years. Not that optical won't grow in terms of its TAM, not that software won't grow, but in terms of the CAGR, we see Next-Generation Metro and Edge significantly outperforming and outpacing the other areas.
As a result, for the past several years, we've been doing a couple of different things. One, we've been undertaking a number of organic growth initiatives. You've seen us come out over the past few years with a set of purpose-built cell site routers for xHaul applications, which we hadn't had in previous cycles. You've seen us talk more heavily about the convergence of IP and optical, you know, and looking at some form of purpose-built switch routers. You've seen us build out our broadband access solution now with the acquisitions of Tibit and Benu. It's a combination of organic and inorganic things that we're doing to really, you know, solidify our position. You know, where if optical were the hunted, you know, in routing and switching, we're the hunter.
We're looking to be the disruptor and the challenger against some very big, well-established players. We're not naive about the opportunity, but we see huge TAM expansion opportunities for us in the area.
What size is that business at this point? What kind of growth rate are you seeing in the orders of that business over the last year?
Routing and switching, next generation at large, routing and switching is about 10, 11% of our business today. It grew year-over-year at a very high clip, almost 50%. That was in part because of the acquisition of Vyatta, the virtual routing software platform they acquired from AT&T. That growth rate won't continue. We do see the growth rate of routing and switching potentially outpacing optical over the next-
20% plus type growth.
Could be. Could be. I mean, maybe not that high, but certainly higher than corporate average.
Okay. One of the other areas that we need to touch on a little bit is the 400 gig ZR as a market, and is that a threat at all to cannibalization of your 400 gig product where you're the market leader? Second, your 400 gig ZR product line and kind of demand you're seeing for it. Could you address whether there's any cannibalization there or any impact on you that you're seeing?
Yeah. Let me just. 400ZR, there's a lot of talk about that and has been for several years. I remember back in 2018, there was a thesis out there that it was gonna cannibalize 50% of our system solution.
Right.
That was five years ago. It has not. 400ZR, I mean, there's a number of different use cases. The primary ones that people think about is data center interconnect, where you've got, you know, high bandwidth capacity requirements, with very specific switch-to-switch interconnect, demands in terms of low power, small space, low latency. What has that's resulted in is a different form factor. In WaveLogic 5, the current, you know, generation of our coherent WaveLogic chipset, we have two instantiations. We have WaveLogic 5 Extreme, which is the purpose-built solution for our telcos, and 5 Nano, which is the power and footprint optimized version for direct data center interconnect.
It is, you know, I think and we think it's best in class in terms of power consumption, all the different other specifications in the market. It's still a relatively small piece of our business. As we've said before, that's not surprising because we think it's a small piece of the overall market. It's a cloud provider play mostly, and that's going to take some time. In terms of your direct question about cannibalizing, to the extent it does cannibalize any system sales going forward, that will be more than offset by the routing and switching and expansion opportunities that we have in that regard. We feel very good about our positioning on 400 ZR and then ZR+ when you start changing ranges and reaches and things like that. We feel very good about our positioning in that space.
I should also point out that there are telco use cases for 400ZR as an on-ramp and for universal aggregation. Again, that'll take much longer to instantiate in the market.
Sure.
It just always does.
Not a lot of discussion about the technology. It used to be, you know, we hear about the 800 gig cycle and the 400 gig cycle. What's the situation in terms of the competitive landscape at 800 gig? Is there, other than Infinera and yourself, is there anybody else that's shipping any meaningful volume, or you guys still own that marketplace, as the, as the cutting-edge leader?
Well, I think we own the market. I mean, with, it's hard to say with, you know, well over 100 customers, over 50,000 ports shipped. I mean, it's a significant market share of single- wavelength, 800G. Infinera is out with another solution, you know, for those operators who want a second source, they're a viable alternative. I do believe Huawei has a solution in market. Your qualification of meaningful volumes, I don't think they're there for a whole bunch of different reasons, but we don't see anybody else in the space.
What about Nokia and NEL? Have they managed to ship anything at 800G, or are they still 600G or lower?
We have not seen it.
Right.
Yeah.
Essentially, it's a two-horse race, and everybody needs to source a supplier, and you're the lead player. It's not just an 800 gig product, right? It's flex coherent, so it ships out at 800 gig if you want it there, but it can be 400 gig, 200 gig, 100 gig, whatever.
I mean, look, the reality is client rate is a function of how far you can carry the signal without regenerating it, and not every operator needs a single- wavelength, 800 gig solution. Not every operator has the infrastructure to support it, and it's expensive to do so. I mean, I think your thesis is right. There's a lot of hero experiments out there even now with things like, you know, 1.2 terabits.
That was gonna be my next question.
Yeah, I thought it might be.
Is there any reason to believe that 1.2 terabits will be any more successful than 600 gig, which was a total flop?
No. I don't think so. I mean, if you want me to be direct, you know, look, I think.
I like direct.
I think 600 gig, you know, looking back, it was always an odd multiplier.
Right.
For the client rate. It really didn't provide a step function up across any number of different specifications above 400 gig. If I'm being candid, you know, I think that there were, you know, a lot of promises made that were under-delivered in terms of its specifications and performance capabilities. Do I think 1.2 is gonna be different? Obviously, now I'm looking into the future. I'm not so sure. I still think a lot of those dynamics play, that I just said about 600, would apply to 1.2 terabits. It is free to put out a press release and put out a slide deck.
I think if you look at Ciena's history over five and soon to be six generations of coherent optical technology, I think the proof is in the pudding and the fact that, frankly, given the current dynamics, our past two years in terms of order growth have been 29% and 26%. The reality, Alex, is our customers are voting with their wallets.
What about Huawei? We don't hear a lot about Huawei these days. But, you know, it's pretty clear that they're challenged at a minimum. And there are some markets like in North America, like in EMEA, and India that have been pushed out. They had a 25% share in India, you have a 20% or so share, and that's a nice opportunity. What's going on with Huawei? In a supply constrained environment, is that giving them some time to continue in the marketplace, and now we're gonna start to see that as the supply improves and lead times come down, that you actually can penetrate that at 20% plus op footprint in China?
I think people are scratching their heads about Huawei. There's certainly been no shortage of ink that's been spilled about them and China more generally over the past several years in terms of the risks with respect to network security and so on and so forth. People are scratching their heads as to why, after all that, Huawei still has 18% of the global optical share ex-China. I think it's a function of a couple of things. One, things don't turn on a dime in our space. Again, these are not wireless base stations in our space that you can just swap out. These are, you know, in some cases, six-foot racks that sit in a central office somewhere that are interconnected with a lot of other pieces of equipment.
I think the reality is Huawei, even before the Trump administration started using legislation and regulations to start impeding their progress from a network security standpoint, China had already basically told the world what they were going to do, which is to look to stand up their own domestic optical supply chain. They're not shy about telling you what they're gonna do. They said it.
As a result, even before those regulations started going into place, they were buying up significant amounts of inventory across the board. I think that has enabled them over the past several years to maintain the ability to ship equipment to customers. I think also, to your point, there are jurisdictions that have effectively excluded, you know, Huawei from new builds and new business. There's been that loss of business they've been able to redirect elsewhere. There are also plenty of countries and operators around the world who don't have the same issues or concerns with Huawei or China, Inc., that the U.S. and Western Europe and India and Japan have. All those things, I think, Alex, have allowed Huawei to continue to survive.
Not just survive, they're still doing $18 billion a year, whatever it is, in revenue in this space. They're certainly still around, but it's hit them pretty hard. You know, I think it is and continues to be a tailwind for Ciena. I don't think it's multi-billions of dollars tomorrow. I think it'll be gradual over the course of time, but we do see it as an opportunity that we have taken advantage of. You know, others are seeing it as well.
Let's stop there and see if there's any questions from people in the room.
Is the composition of the backlog similar to revenues by, you know, telcos, 60%, rest of them are 40%? Or are the hyperscalers non-telco guys over-indexed in the backlog?
Repeat the question, please.
Yeah. The question was, is there a correlation between the composition of our order backlog with revenues? We don't typically disclose the composition of our backlog other than to say, for example, with $4.2 entering the year, that's about $3.3 of hardware and software, the remainder will be services. Most of which will come in during the year, some of which is amortized over multiple years. We don't break out the composition of the hardware and software piece. What I would say, and this is something we've said publicly, is that a lot of that backlog was allocated and will be allocated first in, first out.
The Tier 1 North America service providers plus the cloud providers were ones who got in very early, as compared to some of the international service providers. I do think that if you're gonna look for a composition, it's maybe more from a customer segment standpoint, but we don't break it out by a, from a product standpoint.
Another question? Okay. If there's no other questions from the field, in the meantime, can you talk a little bit about the acquisition of the Tibit and Benu Networks?
Yes. We're super excited about this space. I spoke earlier about, you know, one of the pieces of Next-Generation Metro and Edge being fiber broadband access. I don't think you need me to tell you that there's a significant amount of interest in this space with huge amounts of government stimulus funding in the U.S., and not just in the U.S., but around the world as well, in terms of bridging the digital divide. We do expect that, for example, in North America, more fiber to be laid in the next five years than in the past 15 years combined. You know, we look at our existing broadband access solution, which is robust, in that we have purpose-built cell site routers, a disaggregated operating system, we have ONUs and ONTs.
You know, there were pieces that we were partnering with, we acquired Tibit, who we've been partnering with for several years.
What does Tibit do exactly?
Tibit is a micro plug that fits with the Optical Line Terminal. Basically turns any carrier-grade switch into an Ethernet to PON device. That plugs into the switch at the switching center or the central office.
Roughly the size of Tibit and, you know, their inability to grow is probably because they're too small for most of the service providers?
Oh, I think, yeah. I think.
Now you can leverage that.
I think there's. Yeah, there's a scale play there to, both Tibit and Benu, quite frankly. you know, where you've got operators who are preferring one neck to grab and a vertically integrated fiber broadband access solution across the board. You take the two of those, and we absolutely see a scale opportunity to really leverage the opportunity that is gonna be in fiber broadband access over the next several years. Frankly, accelerating our time to market on the development roadmap as well.
I've given you guys a little bit of time to think of another question. Another one in the audience? If I were to, you know, look at the OpEx side of the equation, which is more of a managerial decision than, you know, what we're able to see externally. What is the strategy for managing OpEx, managing, you know, hiring, things of that sort? You know, can you talk a little bit about how you're viewing your spend?
Sure. I think it's under the heading of, again, control what you can control. We are obviously aware of the macro environment. We're being thoughtful about that as we always are. However, our strategic thesis is that the demand for bandwidth, which has grown and will continue to grow at 30% trumps short-term macroeconomic dynamics. Because of the strength and durability of those fundamental demand drivers, we're investing into the uncertainty, right? We are continuing full steam ahead with our organic growth initiatives. You've seen I'm not gonna be shy about doing acquisitions, strategic partnerships. We're being careful and thoughtful and not spending where we don't have to. We're looking for efficiency with digital transformation, like every global multinational. We're continuing to invest in. We just did an above market merit increase for our people most recently.
Our view is that as things, you know, return over the next few years, not just the snapback from the past three years of pent-up demand, we absolutely expect to continue to grow share, and grow our competitive advantage.
OpEx growth is probably dependent on to what degree you get parts. I know you've guided to 17% growth. If you were to suddenly get better availability and ship more of that backlog, and were to grow 25%, would you then put some investments in or would you hold the OpEx trajectory as it was before you got that incremental benefit?
I think right now, we've guided for $3.25 a quarter, which is a meaningful increase in OpEx year-over-year for all the reasons that I mentioned. Would we be inclined to jump quickly in terms of substantially increasing that OpEx? Probably not. Obviously, that's a decision we'd have to make depending upon the assumption you wanna make about supply further.
It depends on what the orders come in at.
Absolutely.
As well as the supply chain.
Absolutely.
Okay. One of the interesting questions that I think has to be asked of anybody who's in the networking space, optical space is what's the impact of artificial intelligence on the CapEx spend, the infrastructure intensity of the cloud players? When we look at the growth in GPUs, the capacity is going up at 2x-3x that of CPUs. That network deployment requires massive data inputs to make those AI applications run. The boys over at Arista are talking about 10-20x increase in infrastructure required to run some of these massive data center investments, and actually meshing together 125 MW scaled out data centers in order to provide that capability. How does that impact Ciena?
I mean, I know you haven't built it into your model, and you shouldn't.
Thank you. Thank you for saying that.
On the other side of the coin, isn't that gonna drive the traffic growth and which ultimately drives your business?
Look, I understand the theory of the case, and it's not a bad one, right? You know, you mentioned Arista, right? Arista's coming at it from the back end of the data center, right? We're at the front end, right? They're thinking about the compute engine and the AI algorithms that enable the, you know, the bigger flows and the greater bandwidth. On its face, yes, I think that does call for and make a case for coherent optical switching technology in the data center. It's something we're absolutely looking at, and we see an opportunity down the road.
Great. I think we need to end it there. Thanks, everybody, for coming in. The bottom line here is I think this company can beat and raise in an environment where you're expecting estimates to go down. Buy it. It's only at 12x 2024 numbers. It's probably going to 15-18x. It's gonna outperform, and it's my single best idea for 2023.
Thank you, Alex. Appreciate it.
Thank you so much.