Okay, good afternoon, everybody. Thanks for joining the discussion. I'm Rick Schafer, I'm a semiconductor analyst, and I'm joined today by MaxLinear, longtime CFO, I can say now, Steve Litchfield. You know, Steve has a long tech resume. I think he got over to MaxLinear, correct me, but I think it's going on like seven or eight years ago, I think. Prior to that, Steve, a lot of people still know you from almost 20 years at Microsemi. Thanks for being here, Steve, and it's always great to see you, man.
Good to see you. Thanks for having us, Rick.
If it's cool, we'll just dive in. You guys reported just a few weeks ago, with obviously solid numbers, solid upside B-raise. It looks like the company's really turned the corner across all A markets, if I read that correctly. From that call, customer orders picking up and backlogs building, new product wins ramping. It just kind of checked a lot of boxes, I think, to everybody. I was wondering if you could just give a kind of a quick recap of what you're seeing out there today. I know it's only been a few weeks since you reported, but it has been noi sier this year, I think, than usual for a lot of people for lots of reasons. If you could just level set us, that'd be great.
Sure. Yeah, not a problem. Yeah, agreed. I mean, I think we had a, we did have a good quarter. I agree with your last point there that it's a bit noisy out there. I mean, I think we speak to MaxLinear a little bit and talk about the industry as well, but do feel like, you know, finally turned the corner, feel like business is picking up? It feels like it's been a long time coming. I think there's a few things. I mean, there's certainly a kind of recovery underway, but I also am excited and we can talk more about some of the newer products, but we've got some newer products, new products that are driving additional content in kind of existing applications with existing customers, but we're also in some newer areas as well.
I think definitely seeing visibility improve, bookings have improved over the last two or three quarters. That's really encouraging. I think just lead times in general have extended a bit. Customers are kind of getting back into what I'll call a kind of a normal cycle of ordering product rather than having excess inventory. We're well past that, it seems to me. As we look into next year, I mean, 2025 still, I describe it as a bit of a transition year because I feel like things are improving, but it doesn't feel like CapEx has necessarily picked up that much.
I know certainly there's places in AI and the data center that you've seen some of those dynamics, but I would say broad-based semis and I would say a lot of our end markets that we focus on, it's been a bit of a recovery, a little bit of a transition year, really nice growth this year. I think next year as we, you know, that that's likely to continue as visibility for newer product ramps, new design wins, share gains, things like that are happening, that we'll see that continue into 2026.
Yeah. You mentioned carrier CapEx and, you know, obviously infrastructure. I think it's a 30-year revenue. It looked like it was up over 30% last quarter. It sure seems like it's firing, and some, like you said, maybe restock on inventory or something like that. This is for a recovery, but I didn't know if you guys had been a second on that, just talking about between data center and wireless and storage, you know, what those relative contributions look like, what, you know, what lights at the end of the tunnel you're seeing for those segments, because they all, I believe they all seem to be up in the update. I just wanted to maybe see if you can, you know, give a little more color around that.
Yeah, I do. We're seeing, you know, 30, I mean, you quoted the 30%+ growth year- over- year on the infrastructure side. We are seeing that there's two or three drivers, as you mentioned. Data centers is the one that a lot of folks are talking about. We had talked about doing $60 million- $70 million this year in revenue with our PAM4 DSP. That is exciting. That, you know, on or about doubles what we did last year. We're on track to do that. We reiterated that on the call. Excited about this space as 400G data centers transition to 800G. Obviously, we got 1.6T kind of on the horizon that we demonstrated and talked a little bit about at OFC this year. That is kind of the next one coming. 800 G is, I would say, happening in earnest.
We've gotten through a fair amount of calls, moved to production on that product. A little more backend weighted this year, although we had a really good first half as well. That's going extremely, I mean, I think we're just pleased with the progress we made, somewhat of a broad-based customer base. Seeing good traction on that front. You mentioned storage. We had this Panther hardware storage accelerator that we've had for a few years. Developing a new product there. We actually announced last week at FMS and did a keynote presentation with AMD. AMD and us have been working closely together. This is an offload capability. This enables the customers who have a lot of compute challenges to offload some of that, use less cores, be more efficient, be more cost-effective.
We've got a lot of interest out of the enterprise storage players, but then we're also seeing interest from data center, tier one, tier two data centers also on this particular product line. Smaller product line, it's probably $15ish million this year, but I do think this, you know, upwards of maybe $40+ million next year. Can be a more meaningful contributor as we look out in 2026 and beyond as this need continues to go up. The competitive landscape, really, Intel is kind of the main guy. They have a solution called QAT, more of a software solution. Ours is more of a hardware firmware solution that, performance-wise, is quite a bit above that level. I think folks have been really interested in looking at an alternative. I think AMD is getting a little more traction.
Some of those customers are coming our way via them, and then some general concerns around Intel may be, you know, pushing people in our direction. It's a pretty exciting time for that product. The last piece, and, you know, wireless infrastructure has historically been close to half of this product line, or I'm sorry, in market. That also is somewhat of a transition year. You know, 2024 was, you know, CapEx spending was way down, particularly in access. You know, that 5G deployment kind of happened, what, two years ago now. We're starting to see improvements, whether it be inventory easing or some additional spending, but I think that probably continues next year. We've got a couple of new products here too. I mean, we participate in backhaul as well as access. We've always had a modem.
We've added a transceiver over the years, and now we have a DFE. Between the DFE and even, even I would say eBand, like eBand has got more content. ASPs go up. We're seeing a nice pickup in revenues just from content increases or, you know, ASP increases in a particular application or in a particular system. As CapEx starts to return next year, I think, you know, we can add, and hopefully we can even beat the growth that hopefully that we'll attain this year. I think we can maybe do that and maybe a little bit more next year.
That's a lot of great color. Thanks. I might come back to wireless if that's cool, but I'm curious on PAM4 just because it does get a lot of attention. It's a pretty attractive market. Curious if you could help level set us. Like I know you said $60 million-$ 70 million this year. Can you level set versus what is that versus last year's contribution? The second and bigger part of my question is sort of how do you compete against the sort of, there's like two, one's the true 800-pound gorilla in this space and one's the 800-pound gorilla in almost every space, you know. You're competing against two kind of established guys. How do you get in there and win against them?
Sure. Yeah, last year, you know, we didn't give an exact number, but let's call it between $35 million and $40 million last year. A really nice improvement year- over- year, and hopefully we can kind of do the same in the coming years.
Doubling almost.
That's right. That's right. Maybe to your point, I mean, how are we winning? Look, we come to market with a lower power solution. I mean, we hit all the performance metrics. I mean, if you've, those of you who are familiar with MaxLinear, I mean, we've always kind of led with, you know, technical differentiation. Power performance, it really ends up, has what's allowed us to win share. I think it's what ultimately will continue to differentiate us in future generations as well. Yeah, there's one big guy that's kind of dominated the space. Look, the market has gotten a lot bigger. I mean, I know when we started our development probably five or six years ago, this was like a $300 million- $400 million market. We were thinking it was going to be a $100 million product, $100 million product line.
Now clearly that business is, you know, more than quadrupled. I mean, the SAM has quadrupled, right? Which is pretty amazing. That, that's exciting. I think what that's driven is the, you know, customers want to see alternatives. You know, they don't want to be beholden to those two big players, and they've been really pushing for alternatives. I think we've come in with something that's very unique, very differentiated. To add to that differentiation, our next generation solution is being done at Samsung. Like we're currently at TSMC, but we're going to move to Samsung in the next generation. It's another way for us to differentiate from an ecosystem standpoint. They can come with something unique. I mean, all these guys are often worried about supply or you get into export control factors as well.
Being tied to a Samsung ends up being able to kind of help us to stand out even further on a go forward basis.
When you say next gen, are you talking about 1.6T, Rushmore? I'm curious on that, do you stay on a five nanometer? Do you need to be on three to be competitive with some of the other stuff that's out there? I'm kind of curious how you're approaching that market. I know it's not, like you said, it's not doing the heavy lifting this year. That's 800G, but as we move to next year and look to 2027, right? It's going to be a pretty big note.
No, it is a big lift. We are moving to Samsung. We're at five now, moving to, this is kind of Samsung's equivalent. I don't know, it's somewhere between three and four, if you call it that, but it's not the three at TSMC. We think with the architecture differentiation that we have, we're actually going to come through at lower power levels.
When you say it's coming and it will be a heavy lift?
The work's been done. I mean a lot of the heavy lift has been done, right? We have silicon, we demoed this at OFC. We're getting great feedback from customers, and I feel like we're in a good place right now. Is there still work to be done? Is there, you know, tons of work to be done, honestly, from a firmware standpoint? Will there continue to be more efficiencies to be gained here?
Absolutely, there will be. I think we've made just an incredible amount of progress. By the way, we're even looking out further beyond that. That's a typical question that you get. What are you doing 400 G for lane, right? Those are the things that we're engaged with customers on right now, thinking through how to solve those problems. That might be at the translative level, and it might go even beyond that, right? In this domain, these guys really want to work with partners that are able to solve a lot of the problems that they have. I would say our Ethernet signal capability, some of the analog capabilities that we have, is very unique. It does allow us to differentiate, and we got to continue to lead that effort with some technical innovation.
Are there any, you know, I know you said most of the work's been done or the work's been done, but you hear a lot of things, particularly about lasers, as we start trying to push like 200G lasers and stuff. Where do we sit with that? Is that sort of the crux to wider deployment in 2026? Is that too simple of a way to look at it?
No, I mean, I don't think it's too simple, but I guess I'll keep it simple and say that, look, there's still a lot of technical challenges at the system level, lasers included, as you pointed out, that need to be resolved. I mean, I think as you go forward, you get into other CPO opportunities that have to be resolved, but then there's lots of different alternatives as well that I think all of these guys are exploring in order to make the systems more efficient, cost-effective, but also meet the performance criteria that they need to meet. I don't know that there's a clear, call it, I'm not sure that there's a clear winner today on those next generation solutions, but I think you're going to continue to see more and more innovation happen there.
I think it's important for us, as I won't say new because this is our third generation product today, but it's important that we're engaged at every level in driving some of the thought leadership that's needed in the space to differentiate and solve problems for these guys, right? In some of these cases, the two bigger guys, I mean, they're kind of moving towards some custom silicon and solving really, really big problems, yet you recognize that some hyperscalers or even tier one and tier two data center guys, I mean, they have a lot of complex architectural challenges in the system itself that need to be resolved. We want to be that partner that continues to work with them to solve those problems, even when some of their other existing suppliers aren't willing to take that on.
That makes perfect sense. Do you sort of engage on the go-to-market there? Is it the NEOs, the CSPs, the ODMs, or is it sort of all the above? Just trying to think how you create that demand pool.
Yeah, I mean, look, maybe just to focus a little bit on PAM4 DSPs. I mean, so naturally the two big decision makers, I mean, it's the hyperscalers or the data center guy himself that is kind of making the call on the module and often the DSP as well. We're also engaged very closely with the module provider themselves. That's something that I think we have really beefed up over the last few years in those relationships and strengthened those ties. While they may not necessarily be the end decision maker, by working closer with them, now you're optimizing a product offering that really enables them to stand out, to enable them to sell their product over the competition. I think that's something that we've really improved upon over the last couple of years anyway. We think it can be a differentiator as well.
I mean, if you haven't done the work there, and I would say in our first gen solution, we focused more on the hyperscaler rather than on the module guy. Knowing that the decision maker was the hyperscalers, but I think we learn in order to optimize that system, in order for the customer to really see the performance, that, you know, it can be, you got to be tightly coupled on the firmware side, on, you know, even getting into the entire system, including the laser.
Yeah, that makes perfect sense. As part of that, when you're trying to kind of broaden your footprint, which it sounds like that's sort of the way you describe it, basically bringing more value, all that. I mean, you guys had the IP, you had the know-how to, you know, enter new markets like AEC or ACC or things that could kind of broaden that reach. I'm curious, how do you look at those markets? Is that strategic? Is that tactical? Is it like, you know, or is that not of interest? Does it spread you too thin? I'm just curious your approach.
Yeah, it's good to point out because, I mean, one, this market has gotten a lot bigger and kind of the derivations have also grown quite a bit. Public company investors are certainly familiar with different aspects of it. We look at the optical side, we look at the electrical side. We feel like we've got the IP blocks to differentiate and play in either side of that. AECs, AOCs are definitely something that we are working with lots of customers on now, expect to be in production this year as we look at even different architectures, right? LPO gets brought up quite a bit, LRO, so all the IP blocks to support that. All that kind of coming back, Rick, I would say that we also stay really focused on the sweet spot of the market, right? Where is the biggest volume?
Optical transceivers is where the biggest, it's 80% of the DSP market. Yep. That is where our focus has been and where it'll continue to be. Can we and will we continue to diversify as our customers kind of pull us in that direction? Hey, I want to go, we've got customers using LRO today and hey, we want to go in that direction. Okay, no problem. We can support that. Hey, we want to do LPO. Hey, we want to do whatever. We can certainly manage because we have these IP blocks. Really, coming from a CFO standpoint, you really want us to leverage this IP the best that we can. You want to broaden it and address as many markets as you can based off of that same core IP. I think that's something that we've done well and we will continue to focus on.
Yeah, yeah, R&D reuse, right?
That's right. I mean, it's really important. Even if you look further out, you talk about the SerDes capability that we have is very unique. We build our own SerDes. This is something that we can kind of take to the next level, even going all the way to CPO. You need these SerDes blocks there, and it's something that we have that others don't.
Right. No, that's a perfect segue to my CPO question, which is just, it seems like the more, you know, we wrote a white paper on this last summer, I went into a deep dive because I was trying to figure it out. Everybody's talking about CPO. You get down to the nitty-gritty and the rubber meets the road. It seems like even the biggest future players in this space are saying we're years away, like not in the next 12 months or 18 months, but more like three years down the road or something before we see wider adoption. I'm curious about your MaxLinear CPO roadmap. I'm curious what you see as the hurdles there. I mean, we all know cost is out of whack still on it. We all know reliability. Is it the specs that just don't seem to be there yet?
Nobody likes catastrophic failures, so yeah, I'm just curious, sort of where do you think we stand with some of the issues that we have to tackle and where do you guys position yourselves or how do you position yourselves there?
Look, I think you hit the nail on the head. I think it is far out, but I also think that it's really important that, I mean, look, customers are working on this today and they're going to continue to work on it, right? They're going to continue to get closer. We've kind of been through these cycles. I know you've written a lot of white papers on, you know, the last time on LPO, that's not coming. Each one of these, it's like, it doesn't mean it never comes. It may not come right when they think it's going to come, right? I guess I don't want to downplay it from the standpoint that this is where you've got to be engaged with the customer to solve those problems, right? The system kind of gets taken care of over the next few years.
I also think that the industry has gotten so big, so complex that there's no one answer. The investment community often kind of gets bought into like there is only one big answer. This industry, I think, is getting so big that, you know, there's a lot of slivers. CPO can do this in a closed loop system sooner rather than later. Maybe it doesn't take 10 years, right? There's a whole lot of in between. They could carve out a closed end system to address it today. In the interim, they'll start to move towards something that is more mainstream, but it might be four or five years away. I think LPO kind of falls in a similar category. The other thing about the CPO point, and I think you made it to some degree, but it's that, you know, there's going to be these 30s blocks around it.
It's not as if, you know, a player like ourselves or some of the other players either, that just because you're a DSP player today, it doesn't mean you get left out of this entirely, right? There's certainly, you know, capabilities and blocks that are needed. I don't think that changes. Now, are we looking to be a full-on optical guy, you know, and change the entire business model? No, we're not, but I do think that there's a place for us to continue to differentiate and provide value to our customer base.
No, that's clear. I might pivot a little bit, talk about wireless for a second. You mentioned it earlier, you highlighted it. I mean, you know, we carry a CapEx. I mean, for my whole career, it's notoriously lumpy, right? It's just the inside joke on carrier CapEx, but it seems like we're in an uptick on carrier spend. I don't know if you agreed with that, but I was curious if you agree. I'm curious what your thoughts are. This is an impossible question. What your thoughts on duration might be on the technical, you know, kind of a setup for 2026. If there's any regions, are there any regions that stick out to you as either on the plus or minus side that you see?
I think there's a couple of different, I mean, when you say carrier CapEx spend, it's even broader, but I agree with that completely. I do feel like we're starting to get, as I pointed out earlier about the kind of transition year, I don't feel like it's, I think it's improving. It's not great, but it's improving. I think it'll be better next year. I think all that's good, but it's never been a huge growth driver, which is why our strategy has always been to grow content, right? How do we continue to add another block next to, that's why we have the modem and then we added the transceiver, right? You double that share, after the transceiver, we added the DFE, right? Each one of those can get a little bit bigger. We have more of an influence and frankly, just more dollar content.
We've continued to be able to do that. That's how we can, that's why, that's how we can kind of control our own destiny and then navigate some of those ups and downs of CapEx spending altogether. I think it is improving. It's not robust, I wouldn't say. There are places, when you say carrier CapEx spend, PON falls in that category as well. I think you are seeing a lot more dollars flow in that direction than wireless base stations, right? Wireless base stations don't feel like there's a whole lot going on, but even there, I think it's improving geographically. We're exposed globally everywhere, but it is just the bigger markets. Maybe one most folks are familiar with, that wireless infrastructure market.
Maybe the one unique play that we have is backhaul, where you could argue it's a little niche-ier, but at the same time, it allows us to bring a level of differentiation. I don't think it's as exposed to some of these downturns. Where they're running a microwave backhaul system versus fiber or something like that is something that's somewhat unique and it is driven by different geographies. We've seen some pickups in some greenfield applications like in India, but then also it applies very much to Europe, where they're not rolling fiber necessarily between base stations as well. That's something else that I think we can bring that's somewhat unique. Maybe we can, yeah, that probably covers your question. If you want me to kind of go into PON a little bit, we could maybe talk about that off of the carrier spend maybe.
Oh, for sure. I mean, I had one more follow-up on wireless.
Oh yeah, go for it.
Yeah, I got one about PON too. I mean, we kind of look at this as like year seven and these are typically 10-year build cycles for 5G, you know, about year seven or so. By my math and the stuff we're reading and looking at it and forecasting, it's sort of roughly a million macros still going up a year for the next two, three years. I don't know if you agree with that assessment, but the TAM there seems pretty legit for you guys. If I, again, just do a loose math between DFE and baseband and the components you guys, I mean, I don't know if you've ever sort of given a rough idea of what your MaxLinear content is in a macro base station, but I would assume it's sort of up in that $2,000 range or something like that.
I don't know, you can not wash that or talk about that if you don't want to, but I am curious. It just seems like a legit, a legit SAM for you.
It is a sizable SAM for us. I think one, we've made a big step, call it five years ago and did the transceiver. It was a very sizable market specifically for access. We've now taken another big step in bringing a DFE. As you know, Rick, the real, I think the unique aspect of the DFE is a couple of things. The transceiver guys, the two big guys that the whole world knows, aren't necessarily inclined to move in the digital direction. I think as a real mixed signal provider, we're uniquely qualified to have that analog capability as well as that digital capability, build out that DFE. The importance there is that the alternative is a big FPGA that's super expensive and super power hungry, right? By replacing that, I add significantly more content.
Now this, whether this is in a macro or a small cell, whatever it is, there's a big opportunity for us. The market, not as dynamic as data centers, but as long as we continue to grow content in that application with that customer and in a space against the competitive landscape that I think we can win against.
Yeah, no, great color. Back to PON, just because you mentioned it, I think it's pretty interesting to people. I mean, I think your broadband business, correct me, bottomed, I think a year ago. It's been sort of in recovery since. Obviously PON is a standout, but is there any, like, I'd love to hear more about PON, but also anything else sticking out, you know, Wi-Fi, Ethernet, DOCSIS, et cetera, because that's a pretty decent sized basket for you guys in broadband.
Yeah, yeah, yeah. I mean, look, broadband is really made up of PON and cable, but to your point, we've got Ethernet products, we've got Wi-Fi products that really complement that, right? We kind of talk about them all together. If you look at a most recent win, and I'll talk about all the things that you highlighted there, Rick, it's a new win with the, you know, we've won at the top two service providers in North America with our PON solution. It's a space that we didn't have any revenues in three years ago. This year we'll probably do $50 million- $70 million in this business. I think it's on a good trajectory. We're doing a gateway that is running, call it $40- $50. It's multiple chips. It's like six chips. It hits on exactly what you said.
There's a Wi-Fi chip in there, there's an Ethernet chip in there. In some cases, there's a MoCA chip in there, in some cases, but there's a big SoC, which is in that gateway. PON is twice as big as the cable market. This is now, PON is very unique. There's a variety of PON throughout the world, as I know you know very well. There's some high-end stuff in North America and Europe, where we've won a lot of business. Then there's other places in the world, China, for example, that's a much lower-end solution. Some of that stuff we don't chase, like Huawei, for example, does a lot of that business, very low-end, low-cost, not very differentiated. We typically shy away from that.
Where we can lever our capability with the processing, the horsepower that we get there, and then of course the distribution within the home, that Wi-Fi distribution that we talk about, there's an upgrade cycle going on in Wi-Fi right now to Wi-Fi 7. Started on the client side last year or this year, last year, whenever it was. The access points are really starting to move more in earnest, going forward. We'll see a big part of that start in the first half of 2026. Nice thing about that, not dissimilar, you can kind of see a thread through all of these product lines, but there's a content increase, right? Now you've got $11 or $12 worth of Wi-Fi 6 content going up to $14, $15 worth of Wi-Fi content.
It's something that we want to continue to find markets where we can provide more value, more, you know, higher ASPs in a particular application, with more differentiation, of course.
I'll sneak in just because everybody's going to ask, but you know, on, and I know you're limited what you can say about this topic, you've been asked many times, but on Silicon Motion, you know, kind of any status update on the arbitration, you know, timing and resolution.
Yeah, real quick. Nothing's changed on this front. Arbitration in Q4 of this year in Singapore, and hopefully we'll get some resolution Q1 of next year. Once we get that resolution, most folks, I mean, this is maybe for clarity for everyone, we've, you know, terminating the deal with Silicon Motion. We terminated this on MAE. We don't think we owe anything. We've stated that. They have said that we owe the break fee as well as legal fees and damages, which we don't think there's any precedent for. It goes to arbitration panel in Singapore. A lot of folks will think through that. Even if we were to have, if there was a judgment against us, arbitrators are well known for splitting the baby, I say. That piece wouldn't get transacted probably until 2027 or 2028, just kind of the way the dynamics work between Singapore and the U.S.
Looking forward to getting that behind us. Balance sheet, since we're here, Rick, I'll take a bit. Balance sheet, look, we are two quarters ahead of plan on getting operating cash flows back to positive, so we're pleased with that. We've done a lot of cuts on the OPEX front. I think cash flow generation improves from here. We do have a $100 million revolver. Done some acquisitions over the years. Hopefully we'll get the Silicon Motion thing behind us, and then we can start to look at some things down the road. Not in the immediate term. Right now our heads are really focused, heads down, winning more business, getting more market share.
Great. Good luck to you. Like I said, it's always great seeing you and great catching up. Thanks a lot for the time.
Thank you, Rick. Yeah, it's good to see you. Thanks for everybody.