Hi, good afternoon and welcome to our 49th Annual TMC Conference. My name is Bill Peterson, I work on the Semiconductor Equipment research team. We're pleased to have MaxLinear CEO, Kishore Seendripu, and CFO and Chief Strategy Officer, Steven Litchfield. I've asked the team to spend a few minutes providing a quick overview of the company. I will then move to the fireside chat with a number of questions. In addition, please feel free to submit questions using the blue Ask a Question button under the video player, and I'll do my best to incorporate them into the session. Welcome, guys. Glad to have you here, and I look forward to the talk. It's all over to you, Kishore.
Thank you, Bill. It's a pleasure to be at this virtual conference from JPMorgan. MaxLinear, this is our 10th year as a company post our IPO. We specialize in making highly integrated radio frequency analog and mixed-signal digital SoCs. Our target markets today span all markets where there are massive data bottlenecks. What we like to say is that we solve data bottlenecks in the communication networks, from the cloud, over the cloud, into the home, and throughout the home. By that, our markets basically enable us to take our really broadband RF signal technology to solve the data bottlenecks be it inside the home or outside the home. Our biggest part of our revenues today are in connectivity and broadband access.
In that basically, we do the consumer premise equipment where we receive broadband data signals, be it over cable or even fiber, and distribute inside the home using Wi-Fi. G.hn is a power line connectivity and MoCA, which is again broadband connectivity distribution over coaxial cable. A very comprehensive connectivity portfolio, probably one of only two players in the world with such a comprehensive portfolio in connectivity, which includes Ethernet. On the access side, we are the only platform player that provides both the access technology, the gateway processing technology, and all the software that goes with it. 100% of our silicon is on the gateway CPE device, so you could us call as a one-stop shop. We have the same strategy for both cable and for fiber. Fiber is a market we have just entered.
We got some significant market design wins that will show up as paid revenue growth prospects as we exit 2021 into 2022 and beyond. We're very excited about that. The build in this market area for us has come through content increase primarily, where originally we would only provide the access piece, where we receive the signal, but now we do the gateway processing and the distribution as well. Our content has more than tripled since maybe two years ago. On the Wi-Fi connectivity side, we're very proud of our Wi-Fi portfolio. We're one of the few big players with the world's most premier WAV600 product and soon Wi-Fi 7 product line. That's a tri-band solution, but really enabling 10 Gb type of data bandwidth distribution inside the home. We are on the Wi-Fi Alliance, actually.
Anybody who wants to do Wi-Fi has to certainly take this new product. The next big market we have been investing for the last four years is infrastructure. It constitutes primarily network infrastructure on 5G wireless and transport. The other area is basically in optical high-speed interconnects for inside the data center. On the 5G wireless high-speed network infrastructure, we do the massive MIMO radio transceivers for the 5G remote radio heads and active antenna systems, and we also are the only single full system solution provider by transport, wireless transport, which is basically a backhaul, which involves microwave or millimeter wave transport. The optical side, we entered the market kind of late. Late is the wrong word. This is all organic investment. We leapfrogged the PAM4 market and started investing in 400 Gb PAM4, 600 Gb per single lambda solutions.
Basically, you get 400 Gb , four lanes of optical coming inside and the electrical side, there are eight lanes of 50 Gb . Together, you have 400 Gb of optical coming in and 400 Gb of electrical going out. We try to leapfrog our competition, get very advanced stages of qualification for a major operator, web-scale operator, and we're very excited that towards the end of the year, we should start shipping in this market. Likewise, on the 5G side, we got some important design wins with three major carrier operator OEMs. As you all know, 5G has been delayed relative to what the world was expecting, we've got very strong traction. In fact, we started shipping in the first half to an OEM for a North America operator recently. We've also announced an alliance with Facebook for their Evenstar Open RAN alliance.
We would be for the macro base station market, and we're very excited about that. On the wireless backhaul transport market, we are the world's only single system solution provider. We have strong traction, which is growing as the 5G rolls out because outside the U.S., wireless is a pretty dominant transport mechanism for wireless, and we are seeing growth happening to that. If you really step back, four years ago, our infrastructure revenues were practically zero. Today, we have told investors in the last earnings call that we are on target to hit about $130 million of revenues in 2021. By any means to build infrastructure revenue in such a hard market, it's a huge testament to our team's execution capabilities and at the same time, the prospects for our growth in the future because 5G access and optimized interconnects have barely started to take off for us.
We're really excited. The third prong to our whole investment thesis, both organically and inorganically, is we've got a pretty nice sophisticated high-performance analog market, which is primarily industrial interfaces and bridges between various serial transceivers or USB bridges and Ethernet bridges products, and also encompasses our power product line. The power product line is incredibly important because on all our platforms, whether it is on the broadband access and connectivity business or the wireless and optical infrastructure markets, there's a significant amount of power content, and the idea is to own all the BOM around our products. The power product usually is as much of the BOM as the main transceivers. We hope to increase our revenues through proliferation of our power products in wireless infrastructure and in the connectivity and broadband access business. That's the overlay of our product lines.
If you have any clarification, please feel to ask. Let me open up for discussion right now.
Okay, t hat was a terrific overview t hanks for that. Of course, top of mind for all of us is right now, demand clearly seems to be exceeding supply pretty much across the board. Every semiconductor company may have different nuances, and certainly, you're in the same position. Can you just give us your latest update on the supply situation? It feels like really this is going to be with us at least through the remainder of this year, potentially into next year. I think you mentioned it could ease maybe later this year, but potentially go into next year. Some peers are saying that demand is exceeding supply by even 40%. Can you help us understand that demand profile for your products and the differences between your various segments? Connectivity was held back in the first quarter. Broadband appears to be held back a bit currently.
Anything notable between product lines?
Great. It's a very good question, Bill. We have fantastic demand, and I want to characterize our demand hopefully in a very differentiated way relative to what you're hearing about other peer groups, if not even the peer group, the semiconductor business in general. For us, our growth is coming. Hopefully, we are all smart about not double-counting orders. We are very good about parsing through that. Our demand is coming on our broadband connectivity side through actually share gains and increasing the BOM content of our platforms. For us, it's really a secular growth phenomenon that we're driving on, and supply in that sense is very critical because as we gain share, we don't want to lose the momentum if there are supply challenges, right?
At the BOM expansion means that every new platform we're shipping, we're increasing our content, so the amount of silicon we need is much more as well.
That's the broadband access and connectivity portfolio direction here. A big part of it is that for the longest time, people were thinking that the broadband access demand is driven from work from home with the COVID situation. That's not true at all, o ur growth is coming from content increase. At the same time, we're gaining share of major North American operator, and we're getting design wins in fiber as well, which we'll start seeing the benefit of. Thirdly, if you look at the operator spend, everybody's announced intentions to increase their spending on the operator side because there's generally a big increase in bandwidth that our people want to consume. We're benefiting from that. For us, supply constraints are really around increased demand for more new products, okay? Our demand is being crimped.
Our supply is being crimped because primarily this year, we've been fighting battle through on the packaging side, substrates and leadframes. We are hoping that we can ease the situation by the end of the year by bringing multiple suppliers online, leadframe, and substrate packages. So far, we have not had demand issues that are being affected by wafer supply. We got that covered pretty nicely so far, right? You never know what happens in this environment, right? I think we feel good that towards the end of the year, we'll ease the situation. Regarding backlogs, it's a very bold discussion. We have got more backlog for the next 15 months than we know what to do with, so to speak.
We want to make sure we ship to real demand at the endpoints and not count backlog that is being provided because of lead times. It is a very different situation. Secondly, we also see a lot of new products growing and design wins in our connectivity, wired Wi-Fi, Ethernet, moving to next year, that we hope we don't constrain to ship, right? These are all brand-new sockets. We're very excited when our connectivity also should exceed our growth projections pretty strongly if we can meet the supply because there are a lot of new OEMs who want to ship our product. That's the broadband side. On the infrastructure side, it's really new product ramps, right? The new product ramps always have a struggle because forecasting demand is very difficult.
In the initial years, early on in the previous mindset, we would be conservative about how we place orders, right? Let's see the ramp settle in, y ou get some buffer and stuff. Now we have changed our tactic completely, and we are ordering from a more of an annual demand perspective, trying to reserve as much capacity as we can so that we don't thwart our own ramps, right? We need to make sure. The new product ramps in wireless and optical, we're being very careful and watchful to make sure we can ramp. There, the demand, again, is driven by substrate shortages, right? We'll be constrained by that. On our high-performance analog product lines for the industrial multi-market or otherwise, we're not limited overly by supply constraints.
We have that very much under control, in a sense that that market in general, we have a lot of inventory, right? It's in the channel, and we've been able to meet demand very nicely, and hopefully we don't fall into supply constraints at all. We're seeing some pickup in the market as well. Hopefully we can continue to maintain or if we can solve our packaging challenges, we should be able to book the demand and have some ease up in supply towards the end of the year. That's our thinking right now.
Okay. I guess, you mentioned you have that elevated backlog. I know it's only been a few weeks since earnings, how are lead times been trending? Are they still staying? Are there any signs of stabilizing at this stage? Again, it might be a little bit varying by product type.
Obviously there's some variation related to product type and what our own inventory supply chain is and things like that, right? The package type. In general, across the board, the lead times have exceeded from what it was two months ago. Right now, I think they've stabilized, t he lead times are pretty long. They can range from 40 weeks to 50 or 50 odd weeks right now. The situation right now is predicated on packaging and lead frames, right? Substrates. We don't see the effect of any foundry issues yet. If that situation remains that way, we should start seeing easing up going on by the end of the year. At this point, those are the lead times we're dealing with.
Okay. Maybe moving on to broadband and connectivity, kind of grouping them together, although there's a bit different products. Obviously, you had the acquisition last year of the Intel Home Gateway products, which brought a lot of scale and you're now offering more complete solutions. You have Wi-Fi business. I guess just maybe for broadband, what is your current view of the, let's say, the transition from DOCSIS 3.0 to 3.1, and how much more runway do we have with the current transition?
I think there's a lot more room for that transition to happen. I think we are hitting right now more than 50% penetration, driven primarily by the North American markets and some Western European markets. There's opportunity for 3.1 to keep growing as a share as it replaces 3.0 in the other markets over the next two to three years. In general, my own personal experience is that you get closer to 100% over a four-year window, we are now in the beginning of the third year window, we crossed the 50% point in this ramp. I think there's some more runway left, where basically it means good things in terms of ASP expansion, right. There's some more room left here for 3.0 to be replaced by 3.1.
For us, the bigger part is that on a 3.1 platform, we have a BOM extension going on, right? That's even more relevant for us. It's going to be disproportionately more coming from the BOM extension than you would otherwise see from a volume growth.
You discussed market share gains. This is really, I guess, share within your customers as well as maybe your customers' market share gains. How much of your growth expectations from here are based on further share gains versus market growth that we've seen here?
We are pretty much hitting the run rates of where we are typically in the cable market worldwide, right? It's a two-player market, i t's 50% plus minus 5% shares we have always maintained. They may fluctuate on a yearly basis. In the previous two years, we have taken some hits because there were some issues for our major OEM to have qualification a major North American operator. Beyond that, I think you'll get the gain share going another next two to three quarters. By the end of the year, we hope that we're at that sort of equilibrium point, and that's our expectation. You got a couple of more quarters of growth left. With all the supply situation and everything, if you get more market share, you can supply more product, absolutely.
Yeah.
At this point, I would say we got maybe a few more quarters left for the share gains to happen.
Sure.
If we continue to the point, yes.
Understood.
Of course, there'll be other markets where on the fiber side, where we have no share or little or no share, and there we're going to go from maybe 2%-3% of the market to a huge run rate of growth because we have the premium platform for the XGS-PON market. A major operator has selected us in North America, one of the premier operators, and that starts to become a showcase for other operators to adopt our solution.
Okay. Again, it's kind of related, and you talked about BOM increases. The connectivity business, which you guys included your prior MoCA, G.hn, plus newer segments such as Wi-Fi and Ethernet. I guess, can you help us understand the relative size of the major components today? You've obviously talked about Wi-Fi potentially doubling this year, potentially doubling again next year. Try to get a feel for the various growth outlooks for the various segments. First at a high level, and we'll dive in a little bit further.
Let me just get a few words in before I ask Steve to elaborate on this. All our connectivity products are growing. They're all growing very, very healthily, so i t's not just that one. One is a bigger sized one, but they're also when you combine the environment connectivity, they are not small either. Right? Steve, you want to take it from here?
Yeah, Bill. We haven't broken out exactly all the pieces of the connectivity box there, but clearly the largest is Wi-Fi. We did identify that we did on the order of $25 million last year, expecting that to more than double this year. If you sort, as you mentioned on the earnings call as well, that probably tips on to double again. This business, a couple of years ago, Quantenna was running this thing at $200 million plus a year, and there's a clear line of sight kind of getting north of $200 million over the next, call it two to three, four years here, and we're getting great traction thus far with the operators. That's kind of the low-hanging fruit, then we can start to expand beyond that, I mean carriers.
Right.
-outside of that as well with the retail guys also. That's the largest portion of it. Ethernet is a big driver as well, and t hen you do have the MoCA and G.hn business that we've had for a number of years. G.hn is a little smaller business, but MoCA is a sizable business, continues to grow. I think given the whole dynamics that we've seen over the last, call it a year, where within the home, people are really recognizing the need for MoCA, needing to increase the quality of that service within the home. There has been a lot of speculation that Wi-Fi would take that, which with a Wi-Fi offering, that's fine. I think I would look at it a little differently. It's more of a level of market that you're looking for on a high-level product.
You're going to want the MoCA solution, and then on a low-end product, you can get Wi-Fi. Importantly, most of the operators are deploying it with, and so then the user has that capability.
Yeah, you kind of talked about earlier currently supplying in the gateways and so forth with opportunities to expand your Wi-Fi portfolio. Can you remind us of the content opportunities you see for the Wi-Fi 6 boxes and how that is even maybe even better with Wi-Fi 6E? You have the WAV664 series, for example. How do your content opportunities trend for Wi-Fi and then, of course, Wi-Fi 6E?
We have a range of Wi-Fi SKUs and products. If you really look at today, in a major gateway operator box, the Wi-Fi 6 can be anywhere between $10-$15. I would say more closer to the $12-$15 range. When Wi-Fi 7 happens, tri-band, it could be as high as $40+, so $40, $45. You can see that the connectivity product, and if you include Ethernet, if you include a MoCA into it, on any major gateway operator, if they have any of these components, you're looking at this one to be significantly higher than the main processor itself, the network processor itself in the gateway. It's a huge growth opportunity. At the same time, we are not just focused on operator gateways alone. We've actually got major design wins in significant players who are not in the operator space.
For example, you look at major players in China and so on who maybe don't sell the gateway products, but they really love our Wi-Fi products. One of the things was when Wi-Fi 6 happened, they were not focused on an access point offering where you would have double the bandwidth. Traditional ones are 80 MHz and 160 MHz was what the access points needed, and we leapfrogged to that. As a result, there's not a very meaningful access point offering at the bandwidth in Wi-Fi 6 except the MaxLinear silicon. Of course, in Wi-Fi 7, we'd all be on an even playing field on the bandwidth side of it. However, we've got a very, very good position on the Wi-Fi 6 side, basically. Right?
Our 6E is ramping to production now, and then we're working on our Wi-Fi 7 products, so there's significant gains to be had in Wi-Fi in terms of ASP increases. I think you roughly want to think about maybe the gateway platform $12-$15, and then a retail platform will not have that kind of or a mid-tier platform maybe $5- $10. Not $5, but around $10 range, right? That would be the two ranges you want to think about. Any subsequent generation will at least add ASP increases by about 50% or more.
Okay. No, that's great. I know that you guys have been more vocal about it, but obviously one of the pillars you got from the Intel acquisition is Ethernet. You spoke about it a bit earlier. For these same type of boxes, Wi-Fi 6 laptops, industrial laptops, and so forth. How much does Ethernet add in terms of content? I guess maybe collectively, what kind of attach rates are you seeing with Ethernet and really Wi-Fi and what other opportunities do you have beyond the gateway with Ethernet products?
Look, we have a one-to-one attachment on our gateways and the internet product. We have all the generation products that are Gigabit Ethernet, and that are on our platforms. There's a new generation of products where it's 2.5 Gb Ethernet. I just want to point out that we have the world's premier 2.5 Gb Ethernet PHY solution and PHY and controller solution, and we have a range of offerings, a single 2.5. We have a quad PHY, that's four of those and in one. We are very unique in that offering, and we are by and far ahead of any of our competition with a very optimized low-power silicon. We have a lot of design win traction outside of the gateway boxes for even the retail and other people's operating boxes.
One of the things that's happening is the laptops, other products are all going to be migrated to 2.5 Gb. The beauty of 2.5 Gb is that you can use the same Ethernet cabling in your home and everything. You don't have to replace the cabling. Whereas going to 5 Gb or 10 Gb , the cabling has to be completely changed. 2.5 Gb is a sweet spot, which is just about right, where with existing cabling, whether it's enterprise or non-enterprise, where it's a sweet spot appliance. There's a huge growth opportunity in front of us. We're winning designs. It's growing very strongly, and we'll be bringing out more products on the 2.5 Gb side pretty soon.
I would say right now, on the 2.5 Gb side, pretty much we may be the, I hate to say only player. There's competition's product that need to get fixed. If they did 2.5 Gb being thought about, our competition is either selling their 5 Gb or 10 Gb as 2.5 Gb , which means they don't have the cost excellence to fight that, or we are the only product of choice right now.
Okay. That's a great overview of the combined segments. Moving on to infrastructure, this really feels like it's on track to kind of deliver on the promise after we had Huawei bans last year. I mean, you're still fighting supply constraints, but your backhaul business appears to be recovering. Access appears to be taking off a bit. When you look at your growth for this year, I think you said roughly $130 million. Can you break down just at a high level, the main drivers? It seems like backhaul transceivers and modems are a big part of that, followed by massive MIMO transceivers and then maybe just routing and that would be more of the business around the PAM4. If you can kind of break down the growth between those this year, that'd be helpful.
Steve, you want to take that?
Yeah, sure. Yeah, Bill, you're absolutely right. The majority of that, as has been the case over the last year, the backhaul and HBA business are probably the biggest contributors in that infrastructure number. If you recall, the transceiver product for our backhaul market, it's a new product. We've had it designed in for about a year and a half now, and it's just moving to production. We saw a little bit of it last year held back because of the Huawei ban, but we've really seen it come back last quarter. Infrastructure came back really in earnest. A lot of that was driven by backhaul demand, and I expect this new product, as well as our modems, will continue to drive good contribution all year long and good growth all year long in backhaul. That is definitely the biggest.
We saw a big recovery on the power side. Server power management, something that we're very much focused on, expect that to continue to show growth throughout the year. The two exciting areas, PAM4 as well as the 5G access will have nice contributors. The PAM4, I'll hit that one first. To your point, much more back-end loaded, not a huge revenue contributor this year, but much more meaningful in 2022. It'll be nice to finally see that ramp start to happen on the PAM4 side. Lastly, the 5G access piece. 5G access showing good contribution. First product of revenue in Q1 as expected and frankly as planned for some time, and t hat should continue to grow throughout the year. We'll see a bigger contribution each quarter throughout the year.
I think in 2022, as 5G really gets going in earnest, particularly in North America, I think we could see more growth following that.
Sure. I guess, the wireless access, you call this a pretty sizable market, I believe half a billion by 2023. Can you give us a feel for how does this progress from four-by-four all the way up to 64 by 64? I think eight by eight is on the come here. You mentioned earlier O-RAN. It feels like it might be a bit further out, but if you can walk us through the projections out over the next few years of the access business.
I want to point out that in the wireless access, we are not the incumbents. Historically, it's been ADI and TI, and other discrete component makers. We shook up the market by coming with a CMOS 14 nanometer single- chip 4/ 4 transceiver, then the lead product with 8/ 8 RF transceivers. In a sense, we drove the market to these new levels of integration to drive down the cost and sort of help the transition to the higher order massive MIMO configurations. We expect that the transition to higher order, whether it's 64/ 64, is going to take about three to four years to get there. Initial ones are 16/ 16 or 32 / 32. Particularly because China so far has been the driver, and when the U.S. imposed export bans obviously got hurt really bad.
Beyond that, they also changed their strategy and said that Chinese have their own solutions 4/4, and let's stick to that and reduce the order of the massive MIMO to 32/32 so that they've all traded supplies internally. The market has sort of got slowed down and, China, in fact, even now is growing slow, as you must all be aware of it. North America with all the licensed purchases from AT&T and Verizon, and then Dish trying to roll out on their spectrum, we should start seeing a big boost to their rollouts. Now, Dish's will be lower order end deployments because they try to roll out a pretty basic network, and I'm assuming that Verizon and AT&T will go at it, and so will T-Mobile. That will take about three to four years to play out.
I would say they would start at maybe 32 / 32, and then it would head to 64 / 64 over the four-year window. Beyond that integration levels, I would say that it has to rational for suppliers like us to even drive that. Right? I would say that right now eight by eight is a sweet spot from a chip supply point of view. Obviously, moving forward, you could further expect other integration. Now, how does it all tie into our strategy, right? Obviously, that means that moving to more advanced nodes lower the power, because power is a huge part of the 5G problem, because they don't want to bring in extra power, and that's a huge cost of ownership challenges that the 5G players are facing.
For us, the O-RAN development is a long-term game plan where we want to own everything inside the remote radio unit in the macro base. We want to be the platform, just like our gateways. Why not? We have all the technology. We have the transceivers, we have the DFE technology to already do that, and we have the gateway processor technologies. We can put all the pieces together and be the single macro base station remote radio unit. Right? For us, this is a longer-term plan as the O-RAN happens as an alternative to the major OEMs that, in Europe, there are only two, and they want more auction, if you will, and a desegregated system. The interops really become more critical. We want to be right there with our solution. Right?
For us, that's the longer-term game plan is also own the entire platform in an O-RAN world for the macro base stations.
Okay. I wanted to ask you about your 5-nanometer 800G optical product set to sample this quarter. Help us understand, obviously, you're looking to lead product competition at this stage, but help us understand when you expect the products to ramp. Is this 2022 or is this really something down a few years down the road?
Well, we wait for next week, and we may have something to share, right? OFC is around the corner. Obviously, we are seeing some exciting results right now. Let's hold it there. Right? We feel very good. This is about us catching up and leapfrogging the competition. We have the comprehensive portfolio right now, but the big growth is going to come from 400G, 100G standalone solutions, right? The 800G will be the successor product. If you really look at the market today, the hyperscale deployments are pretty fragmented. You have Google and Facebook doing 200G products, and then Amazon wanting to do 400G products now. There's a dichotomy there, and Microsoft sort of behind there, right? They've all agreed that they're all converging on a 400G or 800G platform in the next generation of switches.
You should expect that within a two to three-year window, right? You'll have products ready next year, and then, probably the ramp happens always later than you think, let's say by 2023 end or 2024 beginning, right? What is the difference here between why all of them are agreed upon? Today, if you look at the world, it's 200G optical in for Google and Facebook, and it's 200G electrical out. So four lanes of 50 Gb , four lanes of 50 Gb of electrical, and 400 Gb . The Amazon configuration is four lanes of 100 Gb optical, eight lanes of 50 Gb . Right? So that's 400 Gb in. All of these guys consider the 100 Gb electrical lane as a sweet spot, basically 4/ 100 coming in, 4/ 100 going out, or 8/ 100 coming in, 8/ 100 going out.
The whole world will be harmonized just like 10 Gb . 100 Gb will be the sweetest interconnect speed for almost another 10 years to come and beyond. Just like 10 Gb has been the driver, the whole world will converge on the platform. That's the great thrust of our five-nanometer investment, is to have the best solution for the market where this is going to be the mainstay of the world for a long time to come. Today, with the consolidation of everything and where we are, we can actually be seen in a very good position to be a major player in this market.
That's a terrific overview. Many more questions I could go to. Unfortunately, we're out of time. We look forward to monitoring the progress here of your various product cycles and growth opportunities, and thanks for joining our conference. Really appreciate it.
Thank you very much. You're welcome. Look forward to talking to you guys again. Bye.
Bye.