Good morning. My name is CJ Muse, Semiconductor Equipment Analyst here at Cantor Fitzgerald. Welcome to day two of our Global Tech and Industrial Conference. Very pleased to have MaxLinear with Steve Litchfield, CFO and Chief Strategy Officer as well. Welcome. Great to have you here.
Great. Thank you for having us.
Perfect. Need these.
Gotta get the glasses. Sure.
You know, at a very high level, MaxLinear has been growing its infrastructure business and becoming a more infrastructure-driven story across optical storage and wireless infrastructure over the last seven years. Where are we in that transformation transition?
Yeah. Thanks. Yeah, we have been investing heavily in the kind of data infrastructure. I mean, a lot of products within the data center itself with our PAM4 DSPs, drivers, TIAs. I mean, there's, you know, kind of a broad offering there. So that business is starting to really take off, an exciting time, especially, you know, with what's going on in the market right now. Also got some storage accelerator products, and I'm sure we can talk about those later, but I mean, there's several products on the storage accelerator front, and then some of our wireless infrastructure products that we've been in for a little while. You mentioned the last seven years. Yeah, we've been investing heavily in these markets. It's been growing. Grew 30-ish% last year.
This year, I think some of the analysts have us growing, you know, north of 60% this year. There's a lot of market growth, but I guess I would highlight that we have a lot of newer products that have really, you know, got those design wins last year that are starting to move into production this year.
Perfect. We'll dig into segments deeper later, but perhaps you can start with a high-level state of the union and discuss your views across end market demand as we head into 2026.
Sure. Well, I kinda hit on the infrastructure side.
Yep. Yep.
I mean, that's clearly kind of the biggest driver for the company. On the broadband front, we have some broadband business. I think as I look out over the next couple of years, there's big spending going on, telco spending, CapEx spends, kind of expanding the fiber base. Also a lot of upgrades going on in the cable markets. You know, that has a little bit of seasonality to it. We'll see, you know, that broadband business grew 70% last year. Kinda see some softness in the first part of the year, kinda growing into the back half of the year as DOCSIS picks up. We have a big win at one of the top two North American telcos. That's gonna start to ramp late this quarter on the PON side.
That's pretty exciting. Probably the third one, or you asked about all the end markets, the other one, industrial. Industrial multi-market's been a smaller market for us, but been a tough, you know, couple of years. Finally, we saw that start to pick up in Q4 of last year. We expect that to continue this year.
Perfect. Sure. Maybe digging deeper into infrastructure, grew 30% in 2025, expected to accelerate here in 2026, becoming your largest segment. You know, what are the primary growth drivers, and how do you see your mix, more importantly, evolving over time?
Yeah. Yeah, data center side, I mean, the PAM4 DSP market just continues to go extremely well. Our business, you know, most of the revenues this year, 400 and 800 gig, probably much more weighted toward 800 as it gets more traction in the market. 1.6T is coming. Certainly, we probably see a little bit of revenues late this year, but more of a 2027 production year. So that's probably what the biggest driver is from a data center standpoint.
Maybe drilling a little deeper on the data center side. 800G is the primary driver. How much of that is share gains?
Yeah, I mean, look, this is a new business for us. I mean, we put out kind of guidance for this year. The business has been doubling. Our guidance this year is $110 million-$130 million, so a midpoint of $120 million. So good growth. I mean, most of that is new share for us. So we have been excited. From a differentiation standpoint, the power levels that we bring on 800 gig are about 20% lower than our competition. So that's been the differentiating feature. If you go back to even the technology RF Analog capabilities that we have, that's how, you know, we bring unique performance to our solutions.
Perfect. I think you've publicly talked about major hyperscale customers in both the U.S. and Asia ramping, you know, what does it take to kinda get revenues above and beyond that, you know, $110-$130 target?
We put out that target, certainly, you know, watching the market closely, and we're hoping that we can actually see that accelerate into the back end of the year. Hopefully, we can see some upsides. We spoke to that on our earnings call. What can we do? I mean, you know, certainly we can get more share. We wanna be conservative early in the year on these things. Investors love to get ahead of everyone. Nonetheless, we're working very hard to get additional share, additional wins. The other thing that, you know, we wanna be cautious on these ramps. You know, often customers will shift a quarter or two, and so we wanna be cautious on that, not just to get too far ahead of ourselves.
Makes sense. I think you've talked about Rushmore, your 1.6T product, you know, potentially kind of revenuing as early as late 2026. Curious, when you think about, you know, potential upside drivers, would it be more 800, or 1.6 in your view?
I think for 2026, probably more 800 driven. I think we may see some revenues late in the year, as I had stated, but I do think that there's potential upsides on 800 this year. And we feel confident that we'll see a little bit of revenues on 1.6.
Perfect. I think for Rushmore 1.6 offering you demoed at OFC in April 2025, initial revs 2026, bigger ramp 2027. I guess what differentiates, you know, Rushmore versus Keystone? Do you expect kind of a comparable ramp or might we see even a better kind of acceleration, stronger market share?
Yeah.
from MaxLinear?
Yeah. Look, Rushmore is very exciting. I mean, if I think of kind of some of our history, I mean, our technology has always been that differentiator for us. We've had these great roadmaps, now we're kind of checking the box from a production standpoint. We're in mass production with multiple customers. From a customer standpoint, who's designing in next generation solution, whether it's 1.6 or even, you know, going out to 3.2 and 6.4, you know, they kind of want to see all of those, right? They want to see the technology, they want to see the differentiation, but they also want to see that you've proven yourself, you're in production, with major customers. I think we've kind of checked all those boxes.
1.6 is, I mean, I think we're at the market at exactly the right time. I mean, I'm sure, CJ, you know, I mean, the couple of guys out there, NVIDIA is clearly out with 1.6 right now, but the majority of the market will start in 2027. I think a lot of the market data has it being maybe 20% of the market in 2027, so it's pretty early days. You know, how does that work from a MaxLinear standpoint? Right now, we're really working with customers to get those design wins that'll move into production either late this year or early in 2027.
Perfect. I guess, you know, how do you see kind of the market share, your market share today and how do you see that over time?
Sure. Look, we've had aspirations to be at the 20%-25%. There are two big players in this market. I mean, historically, there's kind of been one guy, and now there's really two players. We feel that we can take 20%-25% of this business. That's what we aspire to do. We've been spending, we've got the product differentiation, so it'll take some time, and naturally, as we progress, we think we can move in that direction.
Is this a market where you're looking to get, you know, maybe second source, or are you looking to get 100% of, you know?
I mean, we'll take sole source business, we'll take second source business. I don't think of it as. I mean, look, the bigger data center guys, I mean, have multiple sources. They have to have interoperability, so that's a common thread. Now, I mean, we've got a few customers where we're kind of the first source, if you will, right? Then it's always kind of fun to be on the other side of that so that they're interoperating with you versus the other way around. Either direction, you know. The bigger ones are gonna have two sources anyway.
Is it for the bigger players always two, or is there opportunity for three?
I mean, often there'll be more. I mean, it kind of depends on the size. It doesn't necessarily mean that. It'll be two or three, and sometimes one, right? It just varies from data center to data center.
Got you. Makes sense. I guess digging a little bit deeper on Rushmore, how do you think about incorporating LRO, retimers, AECs, LPOs, CPO, you know, over time?
Lots of technologies kind of branching off. I think of it, I mean, the market's gotten a lot bigger and will continue to get a lot bigger, right? You're seeing more and more of these applications emerge. I mean, we've been working on AECs for quite some time. We demoed those at the show last year as well. We'll do that again. AECs is, you know, despite some of the rumors out there, it's not dead. It's actually still alive. But even as, you know, joking aside, as we move into 3.2 T, you'll also get into CPO. I mean, certainly there's a lot of talk out there on the LRO front. We demoed LRO products last year as well. You continue to hear more about LRO, LPO.
I mean, I'm kind of talking about all of these from a market standpoint. From a MaxLinear standpoint, look, we have all the tools to do all of these products, right? We're kind of let the customer lead you in that direction, and as the market emerges, then we can kind of chase whichever one. Most of the market today is fully retimed, right? That's, you know, the solution that we have today. As we look forward, if it moves to LRO, then certainly we've got those solutions. LPO, I would argue it's a somewhat of a smaller market, but definitely exists today. That's where your TIAs and drivers, you know, we sell those today alongside of our DSP, so we can certainly participate in that market. And then even, you know, maybe jump ahead to like CPO.
I mean, I'm sure next week everyone will be talking about CPO, if you're not already. That's the next generation. We're working with our customers. I think of it 400 gig per lane. What are you doing there, right? Every customer is kind of attacking a little bit differently. Certainly, CPO is gonna have a place. That's where I think of all these technologies to some degree being disaggregated a bit. Our SerDes technologies that, you know, really drive the DSP today is what ends up being super critical within a CPO application, right? All this kind of gets split out. TIAs and drivers are no longer integrated, now they're standalone. The performance requirements kind of, which I think plays to our favor. We're excited about this.
I'd still say it's early days. I mean, I know investors wanna say that's gonna happen tomorrow, but it's still pretty early days on that front.
Maybe to follow on that, the early day aspect of CPO, you know, when do you think we go into real volume? Is there a way to kinda sub-segment different parts of the market where, you know, you see it first and-
Yeah.
Maybe less so?
I mean, I think there's some slivers, right? I mean, I think you'll see. I mean, NVIDIA's talking about this. I mean, I think you do see some small part of the market, I don't know, maybe less than 20% in 2027. You know, it's probably, I don't know, I think most folks are talking 2028, 2029 before you're gonna see any material usage of CPO.
From a manufacturing perspective for you, how does that change kinda your go-to-market strategy and your foundry providers?
I don't think it changes much. I will hit on use that our Rushmore chip is actually done at Samsung today, which is a unique offering, and our customers have really kind of taken hold of this as they've seen more and more constraints at TSMC. That's something that is unique that, you know, we're bringing to market. As far as having, you know, I mean, we're a fabless company, and we'll continue to use multiple foundries.
Gotcha. You recently announced your first PON design win in the data center.
Mm-hmm.
What is it? What is the size of this opportunity, and how scalable is this architecture across hyperscalers?
Sure. It is kind of interesting. We brought a couple of new products up, or we highlighted them, I'll say, on our last earnings call, and PON was one of them. It's a new area, so we're a PON provider in several other markets. It's only natural that expertise can leverage itself into the data center market. I think it's, you know, hundreds of millions of dollars, probably approaching a billion dollars. You know, it's more of a backplane application. It's not nearly as big as, you know, all this opportunity that we've got on the pluggable side or even the CPO side for that matter.
It is a broader offering that we bring to those data center customers, and as I think of, you know, we wanna be a broader provider of technology to the data center customers, so that was one. We also do some other analog parts, power management parts, serial bridges, USB. I mean, there's a number of things that go into the racks that we're selling in today, so we highlighted that on the earnings call as well.
Perfect. I guess taking a step back, how important is kind of the breadth of product portfolio to kind of penetrating, you know, the overall data center?
It's interesting.
Just one moment please.
Oh, yeah, no problem. Look, I think it's important as a smaller company that we not try to be everything to everybody. I mean, we really do need to stay focused, so I don't wanna imply necessarily that we're trying to do everything from them. But the areas that we have differentiating capabilities, we wanna make sure that we're selling. The Hyperscalers, in particular, are getting bigger and bigger, and so they wanna deal in system solutions and, I mean, there's a lot of verticalization going on. We wanna make sure that we're talking in that language, if you will. Whether you're offering all the products or not, your knowledge and expertise of the system, I think comes in as really valuable to them.
I mean, that's what we probably started this effort maybe two years ago, where we were hiring a lot more system architects, system integration people, in a lot of cases from our customers, because that's the expertise that I think they really rely on. They're not necessarily looking at components for the sake of components. That ends up being valuable to them, and we can help solve problems for them, and that may materialize in a DSP. It could be a retimer. It could be a switch. I mean, it could be anything, right? We've got to make sure that we're utilizing all that capability.
Perfect. Maybe you could speak to your Panther storage accelerator. I think you've discussed the revenues here doubling at least in calendar 2026. Could you kinda speak more to kind of the go-to-market strategy, who your customers are that you have disclosed and, you know, perhaps more on, you know, what the medium-term opportunity is for MaxLinear?
Sure. Our storage accelerator, we call it Panther. It's a newer product for us. It is getting a lot of traction from a numbers point of view. It's not a huge market, but the storage accelerator, this offload capability, ends up saving CPU cores, and so there's a cost advantage. There's a efficiency on the server itself. We see a lot of server customers, enterprise storage customers in particular. You ask about customers, one of the customers that we've announced was Dell, so they're a good example of a user of this technology. From a competitive standpoint, I mean, we have a standalone product. There are other software products out in the market that kind of compress and offload. Intel has a product called QAT. I would say that's probably the primary competitor today.
Our product is superior to that one because it's in hardware, so the compression that it does is way better able to do in software. In some cases, customers that are using Intel CPUs will also use our compression technology or our Panther chip. We've been able to take share there, but one of the other announcements that we made, we've got a reference design and kind of a partnership to some degree with AMD, where we offer this to AMD or AMD customers, I would call it. We have a reference design with them. You can see that on their website. We did a little keynote speech at FMS this year with one of their senior CTO people and talked about you know that capability.
When you've got, call it cold storage that you're offloading, our solution ends up being a far superior solution than what's available for them today. Some of the data center guys are showing a lot of interest in the product as well, so we've got several data center guys that would like to do it also. Naturally, they've got a lot of processing needs, and so it's a good fit for them. Some of the really big Hyperscalers will do it in a custom chip. As I think about the market size, I mean, if you include the data center piece, it's probably north of $1 billion. If you exclude that custom piece of it, I mean, it's still $500 million to $1 billion market opportunity. Maybe lastly, I'll just hit that point on the revenues.
We were expected to do kind of $10-$20 million last year. You said doubling. Yeah. I think it does approach about $40 million this year. Somewhat of a smaller product line for us, but I think it's really getting a lot of attention right now, and I think in 2027, you know, we do have the potential to double that again.
You mentioned cold storage, you mentioned data center as well as Hyperscalers and maybe Hyperscalers going custom. I guess, you know, what is the real target market for you here?
Well, in the short term, the real adoption is on the enterprise storage customers. I mean, I mentioned the Dell example.
Yeah.
That's a good one. The FlashArray guys are probably another example of guys who are using it today. The data center customers are starting to adopt similar technologies. At the high level, like a Hyperscaler, I mentioned some of them are doing in custom, but then you have a lot of other data center customers that are looking for alternatives, I'd say.
I guess in terms of freeing up the CPU, obviously we've heard about kind of real shortages here. You know, is that kind of the principal driver, or is it really more, something else?
I think it is one of the drivers. It's definitely drawn more attention as of late because of that. I mean, the compute needs just continue to go up. I mean, it brings a level of flexibility to them that they haven't had historically.
Perfect. Maybe moving to broadband and connectivity. You know, you talked about declining initially in 2026 due to seasonality and DOCSIS 4.0 transition. You know, how should we think about the timing of the next upgrade cycle, exit rate calendar 2026 and perhaps growth into 2027?
Sure. Yeah, yeah. Look, a lot's happening on this front. As I mentioned, there's a lot of big CapEx spending that's happening, particularly on the telco side, but the cable guys are also upgrading to DOCSIS 4.0 in order to compete with the fiber market as well. We're seeing that right now. We saw a pretty decent, I'll call it, recovery last year. I mean, the business grew at almost 70% in our broadband business. Expect that to moderate a little bit in the first half of this year as the telco customers start to ramp with the new PON business. PON's new for us. Historically, we had done a lot of cable business. About 2.5-3 years ago, we penetrated our first PON customer. We've continued to expand that.
I would expect over the next 2-3 years, you'll probably see the PON business, you know, be larger than cable. We've got the two largest North American telco providers today use our solution. Really excited about that. I think that's been a good evidence of our capability and our knowledge of the space. We're taking advantage of it. Maybe back to the original point. I think the first half is probably a little bit softer as it starts to grow in the back half of the year and some of the DOCSIS upgrades happen. One thing I'd note about the DOCSIS upgrade, it's about a 40% ASP increase.
You know, I know cable market itself, there's no unit volume growth, but these ASPs growing at upwards of 40% are meaningful, and of course, share gains are there as well.
Then maybe on connectivity, you know, roughly 17% of your business. You know, what are the key trends there, key products we should be watching into 2026 and beyond?
Yeah. The connectivity piece, there's probably two main technologies in that. One is Wi-Fi and the other is Ethernet. It grew nicely last year. I think it's expected to do well north of 20% growth again this year. Some of that's driven by Wi-Fi 7 upgrade that's happening. Some of that Wi-Fi is attached to our gateways or our gateway SoCs that we do. We're also branching out, winning other business outside of that, standalone routers, even enterprise Wi-Fi opportunities as well. Those are some of the drivers in the business. The ASP on a Wi-Fi 7, for example, is, you know, 25% higher than a Wi-Fi 6 ASP. The other thing or the other product that I might mention is Ethernet.
Ethernet's, there's a one gig to 2.5 gig upgrade that's going on in the market. Now historically, this Ethernet kind of attaches into our gateway, but we've really branched out beyond that, like into the enterprise space, where you see a lot of Ethernet switches and PHYs at these levels. You also see it in small businesses, you see it in industrial applications. There's more than one billion one gig ports out there that'll be upgraded to 2.5 gig. We're getting nice inroads there against the competition. We see a lot of market share gains on that front.
Perfect. Maybe moving to your financial model. Your target model, 65% gross margins, 35%+ operating margins. In the last quarter, you achieved 60% and 16% respectively. Curious, you know, what is the revenue level? Do we have to think more mix? And is there a timeframe, without putting pressure on you where you think-
No pressure.
You know you can achieve those levels?
Sure, yeah, that's been our target model. I'll maybe talk about gross margin first. I mean, gross margins, you know, we've kind of run in the low 60s%, at you know, similar revenue levels, call it that. That came down a little bit in the past two years with kind of the decline on the top line. As we move forward, top line continues to improve. I'd say the bigger thing that influences that gross margin is the infrastructure or the mix towards infrastructure. Infrastructure runs higher than our corporate average, that is growing faster than our other end markets. We do see this bias upwards on the gross margin side. I think, you know, I'm optimistic that we end this year, you know, starting with a six rather than a five.
Certainly some input costs are going up. At this point, we feel confident that with that mix shift that we can end, you know, hopefully above that 60% level. On the operating margin side, certainly over the last couple of years, kind of during the downturn, we kind of restructured. We did some cuts on the OpEx side. We're getting back to these levels. In fact, that's one of the things I'm very excited about in 2026. Of course, the revenue and the top line growth and the traction we're seeing with our newer products. The leverage in the model is really starting to be more evident, particularly in the back half of the year. That's something that, you know, we've always had in our business.
I mean, multiple times at lower revenue levels as well as higher revenue levels, we've run north of 30% operating margins. I think, call it over the next, I don't wanna put an exact timeframe on it, but, you know, 12, 24 months, you know, we start to get back to that 25%-30% operating margins that we've consistently delivered over the history of the company. I think we can get back to. It doesn't require us getting back to $1 billion at all. I mean, certainly at a much lower level, we can get to those levels.
Perfect. You recently announced a $75 million repurchase program, and I think you bought $20 million in the last quarter, I guess. How are you thinking about the pace of buybacks going forward? How do you balance perhaps maybe a focus on debt, though your debt load is pretty minimal?
Right. Yeah. I mean, we do still have $125 million worth of debt. I don't think we have any intentions to address that. To your point on the buyback, yeah, so kind of mid-quarter in Q4, the board authorized a buyback of $75 million and ended up we did purchase $20 million in the quarter. I think there was, I mean, I think a message in there from the standpoint that I talked about our business over the last couple of years during the down cycle. I think we've shored that up. The profitability is improving. The outlook from a revenue standpoint, a market share standpoint is improving quite a bit.
I think the board, seeing the stock price where it was, wanted to take some actions, to demonstrate the confidence in the business, buy back at a lower level and ultimately also to offset some of the dilution that we naturally have on a year-to-year basis.
Perfect. Which fits very well into the last question. We got a minute left.
All right.
What do you think is underappreciated with the MaxLinear story, particularly as you highlight, you know, the board announcing this buyback, you know, viewing the current stock price as very attractive?
Sure. Look, I mean, I think the biggest thing over the last 12 months is the traction we're getting in infrastructure. I mean, I talk about, I think the analysts have it growing at 60+% this year, but I mean, it's been growing pretty consistently 20%-30% over the last two to three years. This is an organically grown business. It participates in big markets. I mean, that's why we invested here. That's, you know, it's very differentiated. Our technology is kinda what leverages into the space. We really developed some deep relationship with customers. It's got a higher gross margin. We think we're in the right market. We think we've got the right products with the right customers.
I think, you know, I get the impression that there's a bit of a show me part of the story. I think from our perspective, we're just gonna continue to deliver, and we're gonna continue to show you that performance. I think if you look back over the last three years, we've absolutely just hit it out of the park on the infrastructure side, and you should expect to see that from us going forward.
Perfect. Well, thank you for joining us, Steve. Really appreciate it.
Great. Thank you, CJ. Thanks, everybody.