MaxLinear, Inc. (MXL) Earnings Call Transcript & Summary

Good afternoon. My name is Bill Peterson. And we're pleased to have the MaxLinear's team here. We have Kishore Seendripu; and Steve Litchfield, the CEO and CFO, respectively. I've asked the team to spend a few minutes to provide a quick overview of the company, summarize the recent quarter results and also the recently announced acquisition of Intel's Home Gateway platform business. I will then move to the fireside chat with a number of questions, but we'll also bring in questions from the audience throughout the session. If you'd like to ask a question, please click on the Q&A button and type in your question. And with that, Kishore, Steve, welcome.

All right. Thanks, Bill. Thanks for having us. Thanks, JPMorgan. We're excited to be here today. I'll give you a quick overview of our recent quarter as well as the announcement of the acquisition of Intel's Gateway business, and then we're happy to take your questions. So maybe just a brief summary. So we did just report our earnings. Given the uncertainty in the world, we did come in line with our preannounced results. And kind of reflecting back on that Q1, our connected home business actually did a little bit better than expected and went reasonably well. And I would say, kind of despite coronavirus and the like, we are navigating that market well, kind of given the circumstances. Coming into the year, the business had been really tough over the last 12 to 18 months. Starting to see improvements. We were seeing that backlog build early in the quarter. That continued throughout Q1. So backlog going into Q2 as well as backlog building in Q3 has been very good. And so that's encouraging. We really expected to see this business run close to $30 million a quarter for this year. We're a little ahead of that in Q1, which was indeed encouraging. And then our guidance for Q2 was to come back a little bit, but effectively coming in about the $30 million a quarter for those 2 quarters. So that's going well. Our industrial multi-market is probably the one area that we've been more impacted by coronavirus in general. Now, this is a broader, diversified business. A fair amount of it is in Asia, China specifically. And so we've had some exposure. So naturally, we saw a big downturn in Q1. Late in Q1, we saw things start to pick up a little bit, and that's continued. The first 3 to 4 weeks of this quarter, sell-through has been incredibly strong. So we're seeing that -- starting to see that recovery, as you would expect to see China and Asia in general starting to come back to market. We've seen that in our numbers as well. So that piece is encouraging. And then the other area that we've been very focused on is our Infrastructure business. Infrastructure has been really where we've been focused over the last several years. And there's probably 3 areas that I would highlight. One is our backhaul business. Backhaul did expect some seasonal softness in Q1. So not surprising to see infrastructure down a little bit in Q1. But we are encouraged. The outlook there, I think, is very positive on the backhaul side. We've got a couple of OEMs that will ramp this year. And so that seems to be on track kind of despite the COVID issues, but we do expect that to happen. And then our PAM-4 DSP product on the Data Center side, exciting new product area for us. First, 400-gig data center ramping with the largest data center provider in the world, and that's very encouraging. Expecting to see revenues in the second half and more meaningful contribution in 2021. So that's going well. And then the last piece is our massive MIMO business. 5G massive MIMO solution, just getting design wins now. So still fairly early stages, but we'll see a little bit of revenues in the second half of this year and then a meaningful ramp happen in 2021. So those are kind of a super high-level view of our business. Our guidance in Q1 was $62 million at the midpoint of the guide, flat with the previous quarter. We are seeing some impacts, but there's also some positive things happening as well. The work-from-home environment has really highlighted some needs for additional bandwidth in the home. So that's helpful to us. Infrastructure, modest impacts, right? I mean, Infrastructure, these are longer lead times. So we'll see less impact. I mean, we're not -- I keep mentioning to everyone, we're not a restaurant in New York City, so very different business and a very different outlook. Gross margin outlook is very good. I think, again, as Infrastructure grows as a bigger portion of our revenues, you would expect that to grow slightly, but gross margin holding in really well kind of despite the revenue impact. In OpEx, we've remained disciplined. Do expect OpEx to go up a little bit this year, but we've done a good job in the first half of the year of moderating. That's been kind of given the revenue environment. But that will still go up a few million dollars this year. And then the last thing that I would highlight is the Intel acquisition this year, which we're very excited about. This is their Home Gateway division. And it's a business that we're very familiar with. Purchase price is $150 million. We are expecting revenue contribution of about $60 million to $70 million a quarter. We do expect it to close in Q3 of this year, probably early Q3. And really great fit with our business strategically, just kind of tucks right into our connected home group. We're selling alongside Intel's product today. Really the kind of crown jewel there is the Wi-Fi business. And that's what we're most excited about, about the growth opportunities, about the additional servable market that we can address. There are other growth aspects as well, I mean, the Fiber business, the Ethernet business. There's some really substantive technology pieces that we get with this acquisition. And I think we've bought this at a reasonable price, tons of technology, very complementary to what we do today. So I'll stop there and maybe Bill we could have...

That's a great review. Maybe just taking it from there with the Intel acquisition. We discussed the market opportunity doubling, basically doubling your SAM dollar rate to $8 billion and also the opportunity of growing at a double-digit CAGR over the next few years. I presume what's driving that growth is Wi-Fi, but can you characterize the growth opportunities across the different components of the transaction, the Puma SoCs, Ethernet and Wi-Fi?

So Bill, the -- we acquired the Connected Home group of Intel, and it's got -- most people are not familiar with the composition of the business, but it is -- one part of the business is at the Gateway access side. So you can look at it as a network processor chip that basically whatever way you get your data, whether it's fiber or cable, then there's a gateway chip. And then that spits it out to a connectivity distribution inside the home. And until this point in time, the Connected Home group did not have a viable offering in Wi-Fi because Intel while it had one of the front -- leading -- industry-leading clients who had a Wi-Fi, they've been investing last few years in acquiring organically building the Wi-Fi 6. So as of today, on the access point side, the Qualcomm, Broadcom and Intel Wi-Fi access points are golden node for certification for Wi-Fi 6. So what I'm trying to say is it is the state of the art, best-in-class Wi-Fi offering that Intel has. So that's a huge growth opportunity for us because in the existing broadband gateway platforms there is no Wi-Fi yet that is generating revenue. This is an attach opportunity for us. Along with that -- so you can see that the growth in the TAMs comes from the -- primarily from the Wi-Fi attach. And then the other thing is that Intel has invested over the last few years of going outside of cable, investing in 10G fiber PON markets. Because if you look at the evolution of the market, there will be -- cable can continue to increase its data bandwidth capability, but that will eventually be replaced by fiber, let's say, right? And they also invest in that. So you can work with telecom operators like Verizon or CenturyLink or any of these guys to provide fiber-based access. So that's the other growth opportunity because today there is no real fiber revenues, but there are design wins that will lead to fiber revenues. And then there's the Ethernet portfolio that has both PHY technologies, MAC control technologies, 2.5G, 5G and 10GE PHYs, which -- along with the MAC as the controller. So you can actually go and sell these not just in the platform but outside in regular. Through distribution, retail or direct sales, you can sell Ethernet products, right? And if you look at Wi-Fi, it doesn't even have to be attached to our gateways, is also an addressable market, what you call the toolbox market. There could be somebody else's gateway device, but the Wi-Fi device in many cases separate the one that you hook up to. So there is that opportunity as well. And then Wi-Fi has become critically important, even the 5G, as the 5G evolves and Wi-Fi, and if data rates increase inside the home, 5G will be assisted by Wi-Fi as well, even for cell identification in 5G world. And Wi-Fi is being looked at as a critical technology. So that's the value of Wi-Fi there. On the Ethernet side, if you look at our backhaul wireless transport, today we have our own radio transceivers and modems. There's typically somebody else's Ethernet, PHY or MAC. Whether it is 10G or 2.5G, we have an opportunity to increase the BOM there on that front. And also sell that Ethernet to compete against other Ethernet PHY and switch vendors in the 10 gigabit and below kind of market, which is basically done through distribution, retail with the pin-to-pin compatible part or to sell with PC to PC vendors like Dell and all those other companies. So large expansion. So there's a huge portfolio of technology that's come -- that's coming, as a result Intel's investment last few years. Now, they have changed their focus under the new regime, and they want to really -- thought that we would be a good place for that because of our historical partnership with them. So it's a massive opportunity, not just in the gateway side, but it's got implications in infrastructure as well as growing our industrial multi-market segment through distribution and retail sales, basically.

Understood. And I think with Wi-Fi, some of the questions we've been getting, and I think you have on the actual call as well as the -- what kind of content opportunity do you foresee in some of these boxes in the future and Wi-Fi, maybe same-store for Ethernet and some different applications? What type of content per box increases?

It varies. You could have a high-end platform, let's say, a cable platform today, DOCSIS 3.1, and that Wi-Fi is a very rich Wi-Fi, right? Supports up to 10 gigabit sort of capacity and Wi-Fi 6 and 6E can do that. And there, you've got sort of MIMO, right? You can have 2 plus 2 plus 2 in a tri-band solutions, and those could be in the teens in terms of pricing, right? And then the Ethernet, depends, whether you do 2.5G PHYs or 5G. They just think about like maybe $1.5 per unit. And if you sell PHY that could be under -- anywhere between $2.5 to $10 sort of BOM increase. It depends on what you sell with the MAC, PHY and that kind of a thing. So it's a BOM expansion on the high-end platforms, right? And on the low-end side, you could sell a 2 plus 2, and that could be anywhere between -- somewhere between $5 and $10. So the prices range between $5 to $10, the mid-market; $10 to $15 in the high-end market. And there is a market below that, but right now our focus is really on the mid-market -- the sweet point of the market is really in the sort of Access Point/Extender markets, and that's where we want to play in right now to increase the growth rate, okay.

I guess in -- I guess, it's close to a month or so since you announced -- a little over a month, what's been your customer reaction thus far? And then secondly, I guess, what's the status? It seems like the remaining hurdles or remain thing is the discussion with the work councils. I just want to make sure that this transaction is still on track for Q3 close.

So, firstly, obviously, if I said to you anything else other than the fact that customers are very positive, it wouldn't be a right answer. No, I am really serious. We wouldn't get this deal if the customers were not positive because there was no reason why we were one of the preferred acquirer of these assets other than the fact that we partnered with them. But more importantly, Intel cares about customers continuing support and a healthy customer ecosystem. And so customers were as important directing this asset towards MaxLinear as our partnerships. That is the true statement. And regarding the -- it's been a month, but from morning we work through various time zones, so -- because it spans Europe, Asia and stuff. So we are very efficient. We have regular sync-up, and we made great progress. There's no regulatory hurdles right now. And the only thing is that -- the bottleneck is basically us trying to come -- this is an asset purchase, so you have to keep in mind. So we're not buying the entire entities where they are. We only take any employees that we need. So we need to go to work councils. These -- in Europe, there will be design sites in Germany and Austria here. So there, there's EU laws, work councils and -- but those negotiations started, and those could take anywhere between -- we assume in 3 months we'll get it done. But that's why we are saying that third quarter is -- July, August is a good time frame to think about it. And can they go faster? Yes. But I think it will be unrealistic, assumed differently because after we agreed with the work council the government has to -- has a review period to bless it. It's not a regulatory requirement. It's just the bureaucracy of it. So I think it's all on track so far.

Great. I had a question come in from the audience. So with the acquisition of Intel's broadband business and certainty in the road map with the platform under MaxLinear has -- have customer discussions been, I guess, moving towards more supplier diversity outside of the #1 competitor, so not only maybe you get content gains, but potential share gains?

Absolutely. When you have a more consolidated platform, you have advantages in being more -- there is ability to have more price flexibility because you're getting more content to work with, right? And secondly, for the first time, there's a one-to-one complete BOM offering, again, to competition. It also creates -- if the share is split right, both will have the money to invest and invest in the road map that's good for suppliers. And remember, in the fiber, we have zero market share today. So that's going to be a gain. We are having design wins on the 10G PON side. So right now, there is no engagement at the lower end of the fiber market. The focus is on the high-end platforms naturally, right? So we believe we'll win there because we'll be one of the most credible offerings in that space as well. So -- and the Connectivity side we have everything you need, including our own MoCA and powerline -- G.hn wireline connectivity. We'll have the full suite of connectivity. There's a retail market on that as well. We can offer wireline connectivity. So I think we expect it to be a growth business on the access side. We call it low- to mid-single digits, maybe more, but we just want to be cautious, but we see strong growth in the connectivity side of the business. That's the way I would sum it up.

It kind of leads to my question. It appears to me that you either have a turnkey or close to a turnkey solution for -- across these gateways with the SoCs, tuners, transceivers, PMICs. Are there any remaining pieces for the portfolio that the combined entity would need? Or is it -- there is a pretty much a complete solution at this stage?

For the broadband market, it's the full -- we'll also have power, which our competition won't have to offer. It's the full comprehensive offering. And we also can expand BOM further on the Wi-Fi side, for example, by investing in PA technologies as and when we see fit. That is the only missing piece. But that's really not something that we can organically do or require. It's not a big deal, really. So we have all the pieces in play.

Maybe just for the standalone...

There is really, really a full suite -- the full suite, and it has applications outside of the broadband market because it's connectivity. Ultimately, you want to think of the company as solving bandwidth bottlenecks be it from the cloud, into the home, throughout the home. And those -- and the technologies are all common. And we have the full portfolio. Now if you think about in the cloud side, we have a data center investment on the 5G side with the wireless, massive MIMO RF transceivers. Then, we have the wireless backhaul and fronthaul in our millimeter wave, microwave. I think the investor presentation we show off this network in our investor day is complete now, right? There are no missing pieces on the technology or the offerings. It's now about making sure that we can do the self-sustaining growth and that is required. By bringing focus on -- and also, the reality is that the investment cycles are now quite different than what were 10 years ago in any of these markets, right? Because of the consolidation that's happened in the space -- so I think we have every confidence that this business will be a very nice profitable business as well and a growth business.

I can just stop the mic there, but I think there's still 15 minutes to talk about all the other good stuff. So your standalone Connected Home business, you alluded to, is kind of largely played out or playing out in the first half, and you're looking at some visibility into the second half where you're seeing some growth. What gives you confidence. I guess, you talked about backlog, but we know on one hand people's bandwidth requirements are going up, on the other hand you might lose subscribers just broadly. So how do you -- I mean how should we think about the second half under those circumstances?

No. I think we don't have any video business really speaking, right, where you're talking of bandwidth versus losing subscribers, right? So we're really not impacted with that. That's been a long time back. One satellite has virtually disappeared from our revenues. So it's really a data business. We are -- all our businesses are data related. And on those platforms, how do you supply power. It's all about the interconnections now. So I think that the backlog reflects a data business to the extent there is a macro level, there is a broadband demand, we'll benefit from it. It's a question of share split. I think what you saw in the last 2 years is multiple things, right? One is that -- the biggest one is really -- 2 things happened. One is that the U.S. tariffs on China manufacturing affected supply chain shifts, and our biggest customer got acquired by -- Arris got acquired by CommScope. That all gave some turmoil. And in the meanwhile, the operator stopped spending. They cut their OpEx spending -- CapEx spendings by 30% or so in cable, right? And so you have a new normal, and we are seeing growth of the new normal. It is the first time we are seeing nice bookings coming in. We have visibility in the third quarter. And we have not seen that in almost 2, 2.5 years now. So we come into the earnings call and told you, well, we are beneficiaries of this coronavirus and massive demand increase, but we were shy of saying that despite the backlog build out because we want to see it for a quarter and see if it's a sustainable dream. But we're getting very good visibility in the third quarter. So we feel good that maybe we are turning around the corner right now. But we'll just want to wait to declare that, so to speak. But remember they've been working off a new normal over the last 2.5 years, and that's the recovery we're building up. Does it get back to a 30% increase in CapEx spend from operators? I can't speak for that. But our growth is going to come from fiber growth and connectivity growth, BOM expansion in this particular market.

Understood. And are you at this point pretty much crossover levels with the DOCSIS 3.1 over 3.0? Or is still more to come? And I guess the other sort of content adder you have a large customer with your connectivity business for the MoCA, you have the G.hn standard in Europe. I mean, how should we think about the -- those businesses in the second half of the year in addition to it?

All the broadband businesses are showing some good strong -- anything in the broadband-related connectivity is showing some good strong backlog developing. That's not just cable data for Q2, Q3. We are very good at it. But we're also seeing our Q1 on industry multi-market was pretty down, if you recall, and we could not explain what was causing it, but we have seen the bounce back. And the booking and sell-through seems very strong. So we don't feel that it's a channel issue, which some companies have talked about. We have not -- we don't see any gap between our sell-through and sell-in, so to speak. So we feel this backlog in demand it's for real. So I think we're cautiously positive. But one big takeaway message is that the execution has been fantastic. Our wireless products, there's a lot of designing activity going on in 5G; optical, they delayed because it dropped in sanitization, not just us, anybody to do with this large hyperscale data center because of the coronavirus thing. But the backlog is building nicely. But one thing we're doing, I think, nobody else is doing. And this is very important. Even in this dead corona times when you can't have business interactions, we are doing something with our Intel acquisition -- Connected Home acquisition that is trajectory transforming for MaxLinear. So we're not just running the business per se as it is, we're doing transformative things for the business that would really position us when this come out as a really meaningful scale for the company, right? So in that sense, we're kicking on all cylinders right now on the execution side.

With this segmented infrastructure, you have a lot of pillars of growth there. Some of that are -- you're already selling and growth seems to be returning. You have some newer projects as well. You call for mid-teens growth in the current quarter. Based on for visibility, just kind of overarching question that takes into account wired, wireless MIMO business as well, do you think that your overall infrastructure business can grow in 2020? That would imply a pretty strong double-digit growth pretty much throughout the rest of the year. We'll get to each individual component, but if you can kind of give us a feel for the infrastructure business in aggregate?

Yes. I mean so -- look, I don't want to guide quarter-by-quarter. But I think as we look at infrastructure, I mean, I do expect it to grow this year, year-over-year. It -- we really did expect to see a seasonal decline in Q1. I mean, coronavirus, we know has made that worse than originally expected. But it really sets up well for the second half of the year, and we do expect to see some growth in each quarter throughout the year. So, yes, I mean, I think it's probably still, again, a lot of uncertainty out there, but what we see on backlog and kind of the things that are ramping right now, we do expect to see growth. I mean I think it's 5% to maybe 8%, 9% this year.

Overall. So that's going to come off of the back of a strong growth in the second half, basically, right? So you asked would it be a double-digit growth second half. The way we hit those numbers is with that kind of a growth in the second half, basically.

It feels like the current growth is kind of being driven maybe with -- I won't call legacy beyond your backhaul business, modems and transceivers. We've spoken in the past, like how does 5G impact that -- still see the content increase playing out? Is that part of this? What's driving the growth in that segment currently?

Both. I couldn't have answered that question a few weeks ago, but it looks like telecom spend is -- has improved. And therefore, there's a demand picking up on the backhaul side for sure. So whether it's 5G or 4G, it doesn't matter. It looks like if you look at the earnings calls of some of these telecom operators, they are doing good. And I think we are beneficiaries of the same demand. This thing though we saw it a little later. And the second piece is that we have new product offerings in the backhaul space, where we've got a couple of big OEMs where we have already designed in and we got the business, but they were not ramping. I think they're beginning to ramp. Obviously, they are having a much stronger had it not been some of these tariff issues with China and regulatory issues that we have now overcome. But the -- this particular OEM is being cautious, waiting for more clarity on the U.S. position. So we could even see much stronger growth. But right now the growth is predicated on all those mitigating factors taken into account. And -- but we are benefiting from the telecom sort of 4G, 5G growth, whatever is happening. Their numbers are clearly reflecting in some pickup in demand for our own wireless segment. And I just want to correct that they are not legacy businesses at all. They're actually investment businesses, right? I want to make it very clear. Our legacy business really reflects the old satellite and that kind of a stuff, and that's done, right? It's -- and we got some legacy long-term industrial revenues that are part of the high-performance analog or industrial multi-market. But those are products that last for tens of years, right? So it's not like we have any legacy business left in the portfolio.

Moving on to access. You've spoken in the last couple of quarters about now having 2 main customers. You've kind of given us a teaser saying, look out for more announcements here soon in the massive MIMO radio transceivers. Competition, last year, we talked about your traditional competitors, people like ADI and CIEN. I think at that time there's also some discussion about the FPGA guys. It feels to me you guys have some maybe some performance advantages, but how is the competition kind of shaping one year later, one year end? How does the competition look in this segment?

It's very, very good in a way for us, right? The FPGA guys are not considered anymore viable. That's always been the case. No change on them. That's their business model, getting when the prototypes are being done, then it goes to an ASIC model. The OEMs mostly are doing their own vertical ASICs or custom ASICs. And then we are the radio transceiver. We are the ones who came in new when the 5G first phase of designs we are done, which is like the 0.5 generation. The big ramp is -- really is going to happen through the next year onwards. And China is a real ramp on 5G. We're in 5G. And we have outdistanced from our competition with offerings in 14-nanometer technologies, while they are all in 28 or 40 nanometers. Ours is superior offering in power and performance. Though we are not participating in what they are deeming their 5G revenues, the real battle is being set up for the next design phase, but it costs down with new offerings, which is interesting. And there, we have the lead on our competition. And so the lead has increased in many ways, so to speak. And when does it translate into revenue, it's going to take some time. We think middle of next year sometime. Do we get some trickling revenues this year? Yes. But I would look at them as exciting as proof points, but not from a ramp point of view. But the middle of next year, we hope to have ramps going on. And there are 4 major OEMs to work with and 2 or 3 beyond them. And we're working with all of them. Typically, these OEMs will have 2 suppliers, multiple platforms. It takes time to win all the platforms before we get to market share, that's the sort of the penetration angle. We really want to have a major one in Asia, major one in Europe and then build from there, right? Obviously, we are on target, all of them, and we're working with all of them. We feel that they all appreciate our technology. And now the next phase is -- the battle is how does the rest of the guys catch up. All we know is that they're also investing in new technology nodes. But it takes a long time to get into FinFETs to production. The FinFET technologies, we're the only ones there, and we know what it takes. Because we have it in our optical space and the wireless space. And now we're investing in the next advanced technology on FinFETs, right? So -- which we talked about the 5-nanometer test chips, right, where we stand out.

Moving on to the optical interconnect. You talked about having your lead customer. You've also talked about some new opportunities. And while not really new, you've talked about it a little bit, but the single lambda 100-gig. We had Inphi on a couple of sessions ago. They talked about at least their market being around $300 million, $320 million this year, but that also includes 200-gig. So I guess, what kind of market opportunity are you truly competing for? And how much of that is related to the 400-gig from your lead customer versus 100-gig single lambda?

So I would say that we have 2 different stories. 400-gig is right now not shipping in mass volume at all, right? So I would say both of us have 100% share or 0% share. It depends on how you want to look at it. You supply to this module guys, who are going to interop and validation. So I wouldn't consider that a ramp at all. This big, major operator is doing the interop and validation. Coronavirus slowed it down, while they have got more demand, these operator -- this hyperscale data center guy. And so in the other company's case, they are shipping. We never participate in a 200-gig solution, right? We just leave probably in a new entrant into this market, but we are in a great position because of the fact that we're considered in the same breath for the 400-gig solution is proof of that. There we compete with them. But then you've got to keep in mind that the hyperscale data center guys are the ones who go to 400-gig and 800-gig and so on and next forth, right? But those are only 35% to 40% -- 35% of the market. Remaining 60% of the market is the -- there is the Cisco's, all those guys who buy 100-gig -- single lane 100-gig solution. They replace the 25-gig. That's a big market. There, we have the lead pole position. We're the only ones with a customized solution, that's power and footprint into the 100-gig module density requirements. There are lot of design wins we have won. We've got major traction there. And right now, we are really cleaning up those designs. And we're very excited about that. If you just look forward to the MaxLinear, 12 to 18 months, we expect 50% of revenue from 400-gig and from 100-gig, right? And we're already looking ahead to what happens to data center side, the next 800-gig. And we talked of 5-nanometer technologies because we believe, right, the power footprint that's required by 800-gig is not going to be sufficed by 7-nanometer. And the thing is you have to be in something like 5-nanometer, and that's where our investments are focused on, on the next generations of technologies, right? And we feel we'll be the first one to market with a 5-nanometer solution for our 800-gig. We're executing very well on that because we already have a history about the test chip side. So we feel really good that this organically built road map for optical will position us as painful as it has been in terms of fighting the tide on the storytelling, but we have our own story, which is comprehensive on infrastructure, be it wireless or optical or access into the home and the distribution. And that lends itself to building a $1 billion company now with the Intel acquisition part of the portfolio of discussion. And I think we can safely say that from a scale perspective.

We have a few minutes left. There's another question coming in. You're one of the few suppliers and maybe speaking of SAM expansion as well, one of the suppliers that can make these complex DSPs. Also, these sort of PAM-4 based DSPs can used for like 5G mid-haul. Is this scenario you're considering as well and trying to intercept this opportunity?

Absolutely. It is something in the road map, right? But you want to keep in mind that the silicon in the mid-haul or long-haul is very, very small. If you want to be a merchant in silicon TAM, when people talk it is to sell a module, right? So you have to figure out which investment makes sense from a priority perspective, right? The real volume in 5G is in the front-haul, not in the mid-haul, not in the long-haul. And the front-haul still is a 25-gig solution right now. The mid-haul is about 50-gig, and that's like a $10 content in silicon, right? And if you sell the optics, you make some more money out of it, right? So it's a matter of prioritization. We already have all the PAM-4 technologies for a 50-gig and stuff. So it's a matter of sequencing. In fact, it's all opportunities. But, in this large TAM, you got to figure out where is the right place to prioritize the investment. Right now, data center is the big focus. 100-gig is the big focus. 5G is a big focus. And then the mid-haul and backhaul, sort of the front-haul. Front-haul is a bigger volume, but still they're using 25 gigs to come on aggregate. So you think about this massive MIMO in the remote radio unit, you could always 25-gig, let's say, 10 lanes of it, combining to 1 single mid-haul, right? So you have a 10-to-1 reduction in the volume and then the silicon cost also comes down accordingly. So you have to really have an understanding of the scale of the silicon opportunity versus no matter who probably generates in terms of conversation, right? I'm just trying to be sober here on that whole thing. That's not my intention to say. It's not exciting for us. It's just a breadth and span of the offering we can actually bring to bear.

A lot of great opportunities for the team. Unfortunately, we've run out of time. The best of luck wrapping this up and having a much more complete portfolio to address your customer solutions. Take care, guys.

Great. Thank you, Bill.

Thanks, Bill.