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

Greetings, and welcome to the MaxLinear Fourth Quarter 2024 Earnings Conference Call. [Operator Instructions]. As a reminder, this conference is being recorded. It is now my pleasure to introduce your host, Leslie Green, Investor Relations. Thank you, Leslie. You may begin.

Thank you, Alicia, and good afternoon, everyone, and thank you for joining us on today's conference call to discuss MaxLinear's Fourth Quarter 2024 financial results. Today's call is being hosted by Dr. Kishore Seendripu, CEO; and Steve Litchfield, Chief Financial Officer and Chief Corporate Strategy Officer. After our prepared comments, we will take your questions. Our comments today include forward-looking statements within the meaning of applicable securities laws, including statements relating to our guidance for the first quarter of 2025, including revenue, GAAP and non-GAAP gross margin, GAAP and non-GAAP operating expenses, GAAP and non-GAAP interest and other expense and GAAP and non-GAAP diluted share count. In addition, we will make forward-looking statements relating to trends, opportunities, execution of our business plan and potential growth and uncertainties in various products and geographic markets, including, without limitation, statements concerning future financial and operating results, opportunities for revenue and market share across target markets, new products, including the timing of production and launches of such products, demand for the adoption of certain technologies and our total addressable market. These forward-looking statements involve risks and uncertainties, including risks outlined in our Risk Factors section of our recent SEC filings, including our Form 10-K for the year-ended December 31, 2024, which we filed today. Any forward-looking statements are made as of today, and MaxLinear has no obligation to update or revise any forward-looking statements. The Fourth Quarter 2024 earnings release is available in the Investor Relations section of our website at maxlinear.com. In addition, we report certain historical financial metrics, including, but not limited to, gross margin, operating margin, operating expenses and interest and other expenses of both GAAP and non-GAAP basis. We encourage investors to review the detailed reconciliation of our GAAP and non-GAAP presentations in the press release available on our website. We do not provide reconciliation of non-GAAP guidance for future periods because of the inherent uncertainty associated with our ability to project certain future changes, including stock-based compensation and its related tax effects as well as potential impairments. Non-GAAP financial measures discussed today are not meant to be considered in isolation or as a substitute for comparable GAAP financial measures. We are providing this information because management believes it is useful to investors as it reflects how management measures our business. Lastly, this call is being webcast, and the replay will be available on our website for 2 weeks. And now let me turn the call over to Dr. Kishore Seendripu, CEO of MaxLinear. Kishore?

Thank you, Leslie, and good afternoon, everyone. Our Q4 results exceeded the midpoint of our guidance at $92.2 million in revenue, non-GAAP gross margin of 59.1%, along with a significant reduction in operating expenses. We continue to see meaningful improvement in our customer order rates and backlog. These solid signs of recovery, combined with new product traction in strategic growth areas such as high-speed interconnect, PON, Wi-Fi and Ethernet provide us confidence that we can achieve continued growth and improvement in our financial results throughout this year. Looking at our key markets, in infrastructure, the expansion of cloud computing is driving significant growth and design activity around high-speed optical data center connectivity. We made strong progress in '24 with design-win traction and product calls for our 5-nanometer Keystone PAM4 product. We exceeded our revenue targets in '24 and are positioned for exciting growth in '25. As of this month, we have shipped approximately 1 million-plus units total of Keystone products across multiple customers into high-volume opportunities. Our initial design-wins in transceivers, active optical cables and active electrical cables are ramping as expected. We anticipate additional qualification and rollout for 800 gigabit and 1.6 terabit data center applications throughout 2025 and into 2026. The superior power and performance advantages of Keystone continue to be the mainstay and focus for our differentiation even in our next-generation Rushmore family of 200 gigabit per lane PAM4 TIAs and DSPs for 1.6 terabit interconnections. In wireless infrastructure, our wireless 5G access O-RAN single-chip radio unit and our backhaul transceivers and modems are essential for supporting increasing mobile usage and data rates as well as new functionality such as Edge AI. We believe we are positioned strongly for content growth and share gains this year as service provider CapEx spend improves and as our continued design-wins at Tier 1 customers beginning to ramp, particularly in the second half of 2025. Also, within our infrastructure revenues, our Panther III Series hardware storage accelerators are providing exciting incremental growth opportunities. At the 2024 Supercomputing Conference in November, we announced a new software-defined storage solution in partnership with Quanta to address the needs of AI, high-speed computing and other data-intensive applications. This joint solution enables rapid access to massive data sets while enhancing both performance and scalability. As we look ahead, our Panther product is strongly positioned within the data center, enterprise storage applications and at the edge of the network with multiple design-wins with major customers and value-added resellers across key geographies. In Ethernet connectivity, MaxLinear has one of the broadest and most competitive portfolios of 200-gigabit Ethernet switch and PHY products for the enterprise and small and medium business switch markets. Swan Creek, our single-chip integrated 8-port PHY and switch is gaining traction across multiple enterprise customers looking to upgrade their networks to 2.5 gigabit Ethernet rates. Our Tier 1 North American enterprise OEM customer is expected to ramp to production in 2025 and contribute to significant Ethernet revenue growth over the coming years. In addition, we are seeing widespread interest from next-generation broadband gateways and routers. Shifting to broadband and Wi-Fi connectivity, our considerable focus on PON to expand our broadband target addressable market in addition to a strong cable data offering of DOCSIS 3.1, Ultra DOCSIS and DOCSIS 4.0 solutions is bearing fruit. We have exciting design-win traction for our single-chip integrated fiber PON and 10-gigabit processor gateway SOC plus triband Wi-Fi-7 single-chip platform solution. We have a promising engagement at another Tier 1 North America carrier, which we believe can become a major opportunity for us in '25 and '26. In conclusion, our strong product road map execution has begun to deliver meaningful traction and target addressable market expansion across several high-value categories, including high-speed interconnect for data center, enterprise Ethernet and storage accelerators, wireless infrastructure, multi-gigabit PON broadband access and Wi-Fi connectivity solutions. As we begin 2025, we're not only energized with the solid growth prospects in '25, but we also feel confident of achieving sustained revenue growth in the coming years. With that, let me now turn the call over to Steve Litchfield, our Chief Financial Officer and Chief Corporate Strategy Officer.

Thank you, Kishore. Total revenue for the fourth quarter was $92.2 million, up 14% from $81.1 million in the previous quarter. Infrastructure revenue was $27 million, broadband revenue for the fourth quarter was $29 million, connectivity revenue was $20 million, and our industrial multimarket revenue was $16 million. GAAP and non-GAAP gross margins for the fourth quarter were approximately 55.6% and 59.1% of revenue. The delta between GAAP and non-GAAP gross margin in the fourth quarter was primarily driven by $3 million of acquisition-related intangible asset amortization. Fourth quarter GAAP operating expenses were $92.4 million and non-GAAP operating expenses were $61.3 million. The delta between GAAP and non-GAAP operating expenses was primarily due to stock-based compensation and performance-based equity accruals of $20.4 million combined, acquisition-related costs of $7.3 million and restructuring costs of $3.1 million. GAAP and non-GAAP loss from operations for Q4 2024 was 45% and 7% of net revenue. GAAP and non-GAAP interest and other income during the quarter was $351,000 and $677,000, respectively. In Q4, the cash flow used in operating activities was approximately $28 million. We exited Q4 of 2024 with approximately $120 million in cash, cash equivalents, and restricted cash. Our day's sales outstanding was up in the fourth quarter to approximately 85 days. Our gross inventory was down versus the previous quarter as we continue to make improvements with inventory turns slightly less than 1. This concludes the discussion of our Q4 financial results. With that, let's turn to our guidance for Q1 of 2025. We currently expect revenue in the first quarter of 2025 to be between $85 million and $105 million. Looking at Q1 by end market, we expect broadband and infrastructure to be up. Connectivity is expected to be approximately flat and industrial multimarket is expected to be down. We expect first quarter GAAP gross margin to be approximately 54.5% to 57.5% and non-GAAP gross margin to be in the range of 57.5% and 60.5% of revenue. We expect Q1 2025 GAAP operating expenses to be in the range of $93 million to $99 million. We expect Q1 2025 non-GAAP operating expenses to be in the range of $56 million to $62 million. We expect our Q1 GAAP and non-GAAP interest and other expenses each to be in the range of approximately $1 million to $2 million. We expect a $2.7 million tax expense on a GAAP basis and a non-GAAP tax of 0. We expect our Q1 GAAP and non-GAAP diluted share count to be approximately 85.5 million each. In closing, another quarter of improvement in customer orders and continued new product traction give us confidence that we are entering our next stage of growth in 2025. We're excited that our innovation and investment in strategic applications such as optical high-speed interconnects, wireless infrastructure, storage, Ethernet, Wi-Fi, and Fiber broadband access gateways are beginning to deliver tangible opportunities for near-term and long-term growth. In addition, our strong focus on operational efficiency in 2024 is positioning us for positive leverage in our business model and a return to profitability this year. With that, we'd like to open up the call for questions.

[Operator Instructions] Our first question comes from the line of Tore Svanberg with Stifel.

My first question is, so obviously, you're still relatively an entrant into the optical interconnect market, but you are the first one to report among your peer group. So I was just hoping that both Kishore and Steve, could you just talk a little bit about the events of this week, especially from Monday, how do you view this whole topic as far as potentially impacting the optical interconnect market?

Well, Tore, we're pretty excited that we now have recorded a strong 2024. We exceeded our own internal targets of revenues. We have design-wins and shipments in various quantities and stages with all the top module makers in the world, with their end customers spanning across both China and the U.S. So that's the exciting part. With regard to what happened this week, I'm afraid that's a question that really only democratizes and really expands the real possibility for new entrants like us to really expand our share as the market grows. And at the end of the day, we are a high-speed interconnect PHY transport company. So no matter what happens in the compute, the links are going to be more and they're going to be faster and speedier and we have the right technology for low power and high-efficiency performance targets that these markets will require. So from my point of view, processing content is one thing, but the links are a given and they're going to be really needed. So I think it really democratizes for people like us to really, really participate in majorly what I call expanded and singularized market if that's going to take hold.

Yes. That's a great perspective. And as my follow-up, Steve, the DSOs have been all over the place this year. They came down very nicely in Q3, but now they came back up again in Q4. Obviously, still not as bad as Q1, right? But help us understand what's going on there? And how should we think about DSOs here into '25?

Yes. I think they were probably understated a little bit in Q3. It was really a product mix and just some of the sales that we had there. So they did come up a little bit, but I would argue that this 80% to 85% range is probably where you'd likely see it the rest of the year.

Got it. But that explains why I think your cash balance came down, was it $30 million? Was that the main...

Yes. The cash balance did come down, but that was already reflected. We talked about that last quarter. There were some restructuring costs, but there was definitely a kind of movement around the balance sheet. But certainly, that was as expected in Q4.

Our next question comes from the line of Ross Seymore with Deutsche Bank.

I wanted to talk about the bookings and the backlog. It's, I think, the third quarter in a row you've seen those improve. So it's nice to see that corner being turned. Could you give a little bit of color on where you see that happening, either by end market or geography? And perhaps even is it company specific? Or is it just kind of the cycle is finally turning? Any sort of color like that on those metrics would be great.

Sure, Ross. Yes, thanks for the question. I mean, yes, we're definitely continuing to see really nice improvement on the bookings side, starting backlog, I mean, going into this quarter versus the previous 2 or 3 quarters is much higher. So we're feeling much more confident. Just visibility has certainly improved. Customers are starting to understand lead times and recognize that they can't get the product when they ask for tomorrow. So we're seeing some nice improvements as far as visibility and forecasting. I mean your question is it MaxLinear versus everyone? It's I mean, look, in the broadband market, it's certainly been a lot worse over the last 2 years. And so I would say that is starting to normalize a bit. I also -- Kishore in a lot of his statements talked about the new products. And I think as we look at 2025, I mean, certainly, we've got a recovery. I think that will be a nice tailwind. But I think what we're most excited about is these new products that are coming and these new programs that we've won. So there's certainly market share gains that we're seeing and new products that are starting to ramp.

Great. And I guess, Steve, probably this one is for you as well. Just on the restructuring efforts, not the fourth quarter impact per se, but I just wanted to make sure that there's no big change. And I think you guys talked about what, $220 million, plus or minus for the full year on the OpEx side. Just wanted to see, if that's still the right trajectory and if there's any sort of lumpiness to the path on that between 1Q and 4Q.

Yes. I mean, so the restructuring is certainly underway and the biggest portion of that was in Q3, and we had talked about kind of some residual that continues after that. But we're definitely seeing nice improvements in OpEx spending. I don't think the expectations have changed. I do expect it to be somewhere between $220 million and $225 million for the year. And I would expect as kind of some of these final effects take place that you'll kind of see it move down throughout the year, modestly. It's not a huge change there, but it will come down a little bit throughout the year.

Our next question comes from the line of Timothy Savageaux with Northland Capital Markets.

Question on the optical front, I guess. I think you've been talking throughout the year and you'd increased this range to more than $30 million in revenue for the year. I wonder if you could tell us kind of where that ended up coming in. And given your volume comments, it seems like you must have a pretty good January, I guess. So can we infer kind of a big step up there? I know you've guided infrastructure higher, just looking at the 1 million unit volume versus whatever your revenue may have been in '24. And I'll follow up from there.

Sure. Tim, yes, I mean, I think I'd probably echo what was said in the previous statements. I mean I think we're really pleased. We came in higher than what we had expected. We didn't -- we're not breaking out the exact number, but we started the year between 10 and 30. And I think we probably landed shy of the 40 number that we were stretching to, but certainly well above the high end of the range that we said originally. So we're pleased with the progress. I think as we've talked about, I mean, a lot of our wins and future production revenues are really driven by 800-gig conversion, and those are just now happening this year. So it's an exciting time and looking forward to talking more about that in the coming quarters.

Or perhaps on the next question, I wanted to see if you could -- if you would like to set a range for this year similar to what you -- not a similar range in the numbers, of course, but conceptually similar about what you think that optical business might be able to generate as you sit here in early '25?

Yes. Sure, Tim. Look, I don't think much has changed on this front. I think -- I mean, we've talked about kind of the $60 million to $70 million number. I think that's a very reasonable number that you can target right here. Hopefully, these new data centers roll out as expected. Some of that's out of our control. But certainly, we're doing our part getting the wins and getting the qualifications completed. And so as soon as -- then we see our customer, I guess, in this case, our customers' customers roll out these programs.

Okay. Last one for me. I guess a little while ago, we saw an agreement -- well, actually, Amazon has been making a couple of agreements and announcements of late, but one with Jabil, who I think is a module partner of yours dating back to OFC, where Amazon took a bunch of warrants and Jabil doesn't exactly indicate an 800-gig transceiver module relationship, but it seems like it could. I wonder if you have any comment on the potential impact of that agreement on MaxLinear.

Well, look, I'm not going to comment on those agreements. I will confirm, as I think many of you have seen, we did demonstrate and we've been working with those guys for some time. But we definitely -- I mean, we demonstrated this at OFC last year, and they've been a good partner.

Our next question comes from the line of Quinn Bolton with Needham & Company.

Steve, I just wanted to follow -- sorry, ask a question to you, sort of a follow-up on Tore's question. Cash down to about $120 million guidance probably has you at a small few million, maybe mid-single digit million net income -- non-GAAP net income loss in the March quarter. Are there any major changes in working capital or any residual restructuring cash charges that hit you in the March quarter? I just -- any sense where you think cash might -- what cash might do through the quarter?

Yes. So good question, Quinn. I think as we've talked about cash, nothing's really changed on that front. We do expect inventory to continue to come down. So that's good. Working capital, I'm sure from -- as revenues start to recover here and we burn down those inventories, we'll certainly have to replenish that. Out quarter revenues are certainly going to be above where they are today. So we'll have to start building on that, and we're doing our best to manage it, of course. We've talked about cash flow breakeven somewhere kind of midyear. And it's probably Q2, Q3, likely Q3 is where I'd probably put it today, but we feel very comfortable with that, yes.

Okay. And then I guess maybe for Kishore, 2 questions. You talked about some design-wins, a Tier 1 design-win, I think, in the wireless infrastructure market beginning to ramp in the second half of '25. I wasn't sure if that was for the Sierra product or if that was for the backhaul product. If you could provide any more detail there, that would be great. And then sort of similarly, I think you mentioned now a promising engagement with an additional North American Tier 1 on the PON side. I think in the past, you've already talked about working with 2 of the largest in North America. So wondering if that is a third Tier 1 in North America or maybe I misheard something.

Well, first, to answer your question of the ramp on the wireless infrastructure in the second half, there are 2 parts to it. There are -- they're both in our wireless backhaul transport business and also on the access side. And so it's for both the product lines, the one for our Sierra O-RAN product and as well as our backhaul products. And primarily, this growth is coming through content expansion of these customers on the backhaul side. And obviously, our Sierra product is unique and is a leader. And it's the beginning of what would be a whole industry trend for merchant silicon to support both macro 5G and massive MIMO revenues in the future. So that design is a little bit of a timing uncertainty. Hopefully, you're at Mobile World Congress, and you'll be able to catch a lot more glimpse of these products that would -- that I'm super excited about for sure. Regarding on the broadband side, a Tier 1 operator, we referred to where we're in engagement. We expect that to be a major driver in 2026. But could it have a little bit of revenue in '25? Sure. Now we talked about 2 major North America operators, but you must understand that we never talked about a gateway design per se in a Tier 1 operator. We talked about there are multiple product lines within these Tier 1 operators, but this will be a whole gateway design-win at a Tier 1 operator.

Is that PON and Wi-Fi or just the PON chip?

Yes, it's both, right. When I refer to gateway these days, it's de facto. It has Wi-Fi. It's a full PON 10G XGS-PON gateway with processor that supports 10-gigabit speeds plus the world's first tri-band single-chip Wi-Fi access point solution, along with our own Ethernet Quad Port 2.5 gigabit Ethernet PHYs.

[Operator Instructions] Our next question comes from the line of Suji Desilva with ROTH Capital Partners.

Just wanted to, Steve, maybe double-click down on the 1Q guidance. I appreciate the segment color, but I'm just trying to understand if infrastructure is likely growing and continue to ramp up here, where the offsets to that are? And what would have you be at the low end if you have a segment like infrastructure that's ramping as strong? Just trying to understand some color as to the puts and takes there.

Yes. Look, I think we're excited. Infrastructure is probably the biggest grower for the year, and that's certainly the one that we've got a lot happening around. But I mean, we'll definitely see broadband grow as we stated in Q1 and likely to end the year at a much higher level. I mean the one that's been weak has been industrial. And I think we're still, like many of our peers, kind of working through that. Demand is soft. There's a little bit of inventory out there, but I think it's really more about demand. And certainly, connectivity is starting to recover as well as we talked about along with broadband.

Okay. That's helpful, Steve. And then maybe for Kishore, I know Tore asked about the news this week, but maybe topically also CPOs being discussed, whether it's NVIDIA's in-house solution or merchant vendors like Marvell. I'm just curious if you could update us on your thoughts on what -- whether CPOs impact to your opportunities orthogonal or whether it creates opportunity or whether it's something a risk? Any color there would help as people are looking for that in 800, 1.6T.

Well, CPOs have now done their third incarnation in the discussion in the data center optical interconnect space, right. But there is always other vendors with optics and CPO is a part of the discussion engagements we always constantly have. But our goal is a pure DSP, PAM4 or TI type vendor story, whether it's a linear optical transceivers, if you will, or the next-generation 400G/lambda silicon for that, that supports it. I look at our presence and focus on data center is beyond optics. And that's why we call that High-Speed Interconnects, right? Because we're a silicon provider, it spans not just the interconnect connection, but any kind of high-throughput interconnects within a compute or storage environment is a target addressable market for us. So CPUs is one element of it. For that, you have to have your own optics and or you partner with optics players to enable that. I just don't think that there's many, many years before CPOs, if ever become viable because of their various issues of quality, yield, power and the footprint and then how they lock in a lot of ASP on the quality front that it goes to waste if it is not really properly actualized. So I think these topics come and go in our industry. And every time people provide different solutions or venture into different ideas, we have to pursue all these directions, okay?

Our next question comes from the line of Karl Ackerman with BNP Paribas.

This is Sam Feldman on for Karl Ackerman. So you indicated that your DSP business will ramp in 2025, given your engagements at hyperscalers. What gives you confidence it can double, given Amazon indicated that 800G may not ramp until 2026? Does this mean the DSP ramp with 800G?

Look, there's a large, large opportunity that is 400G and 800G will again last for many, many years. I think that also answers Suji's question about CPOs, honestly. So our revenues are a mix of both 400G and 800G. And as Tore pointed out, we are new entrants in this market space, and the third player. So we have a large revenue in the optical interconnect that we can go and access. Regarding, there are 2 markets here. We have always maintained that the line side is really delayed relative to what I call the compute side, which is the AI network and the very, very different markets. So the line side market are indeed delayed -- not delayed. I would say they've always been sort of the 800G comes much later. And I think they are on track on the front. And Amazon is not the only one, right? There's Meta, there's Microsoft and everybody else. And our road map, frankly, followed the cadence of the line side markets. And we've always -- we have said that we are not a player in the NVIDIA market in terms of as being what happened in the past. And so that is the reality of it. So I think that you are sort of comparing both the markets in very different markets in terms of the time line, how they're evolving. And same will be true for 1.6 terabit, by the way. I would say that it will be a long while before 1.6 terabit or 200 G/lane that becomes a meaningful portion of the shipment of the revenues until after 400G and 800G have really, really run their course. Okay.

Our next question comes from the line of David Williams with Benchmark Company.

I guess maybe first, and Kishore, I think you mentioned this in your script earlier, and I may have missed it, but I just wanted to see if you could give us a little indication on the 2.5Gb Ethernet PHY product Swan Creek there. And I know you've talked about design-wins there in major Tier 1 enterprise OEM customers with multiple design-wins. But how is that ramping? And maybe just any of the color around the feedback or demand trends that you're seeing for the Swan Creek product line?

So I think Swan Creek, as a product goes, probably one of the most successful products that we have ever designed and the kind of demand for the product. It's premier, it's highly differentiated. It has outlined anybody's offering on the 2.5Gb multiport switch category. And we, frankly, ourselves were quite surprised with the amount of traction it has. And it's pretty much designed with all the major players on routers and gateways and even on the industrial side as well. It's a very unique product. It can do multiple ports all the way from 4 to 8 and 2 of those switch ports can be compounded to do a 32-port solution as well. So it's got extremely good traction. And with this Tier 1 OEM, we were supposed to have actually ramped stronger towards the end of last year. That has not happened, but it's gotten delayed and -- but the ramp continues in a sense that the plans remain intact. It's their major platforms. And I believe that as we head towards the rest of the year, it will start ramping. And in '26, '27, '28, '29, it will be in a pretty strong run rate position. And the overall product really, we believe, can be $100 million per year revenue product line for Ethernet over the next 2 to 3 years. And the mix would be single PHYs and multiport PHYs and switches in equal proportion or more tilted towards the multiport PHY and switch.

Maybe anything regionally that you're seeing to speak of in terms of demand trends around maybe China or even North America? How are you seeing, I guess, geographically, how the demand trends? Anything you would point to there?

Absolutely. Most of the -- almost all of these designs in these markets happen in Taiwan or China. And that's where most of our activities, our support activity, sales activity as such. So I would say that, well, the end markets are quite varied, but primarily the end markets are U.S. and China-centric, which you should expect given they are the largest markets in the world, right? So we're quite happy about it.

Our next question comes from the line of Alek Valero with Loop Capital Markets.

This is Alek on for Ananda. I have 2 quick questions. So my first question is, as we go from 800G to 3.2T, do you guys see yourself becoming -- do you guys see yourself as being more attractive to customers? If so, what do those dynamics look like?

Look, really, really speaking, there are -- you have the entrenched incumbents as one would duly give them credit for, which is both Marvell and Broadcom. They come at it very, very differently in terms of the competitive force, if you will. However, there's only one credible new entrant on the optical transceiver space by far. There's nobody even close to what we offer. And what we offer is extremely low power in the 800G solution and 400G solution for 100G per lane design, which is the latest -- which is the newest generation of products that are ramping or will be ramping soon, like one of the analysts brought about the Amazon delay, for example. So the differentiation really comes from extremely low power, and we all now know whether it's an AI network or any data center, power, power, power is the key, and that's where we built our core competencies as a company. So I believe that -- and we've also seen shortages on DSPs and optical module solutions in the last few years as all data centers try to upgrade to new technologies. So there's a genuine demand for a third supplier. And it is where everybody wants a third supplier, and we hope to first build our position as a third supplier and then build from there. And that's been our game plan from day 1. And the fact that we went to $0 million to $40 million last year is proof of that. It's 1 million units, that's pretty substantial. And hopefully, we could do much, much better this year and leading up to next year. And I think we are very pleased with the progress. Now this has entailed a lot of investment on our side. I know and analysts, you always have OpEx questions, but I just want to be very, very clear that this is a strategic area of interest, investment, a high-growth market area, and we intend to continue to invest very, very strongly in this space to expand our portfolio beyond the optical space. Okay. I think that's where our strategic thrust is right now in the infrastructure space.

Got it. Just for a quick follow-up. So with you guys [indiscernible] in the world of co-packaged optics, do you guys believe that advantages your guys -- disadvantages you guys? Or is it net neutral to you?

I would say it may be a net positive because if our competitors are invested in optics and they try to do a fully integrated CPU, there are more optics producers in the world who are more -- who are competent and really, really excellent at that. And the market capacity will require that those optics players are part of the supply ecosystem. And therefore, they need a pure-play silicon vendor and MaxLinear absolutely is a pure-play silicon player in this. No cables, no optics, and that sort of a thing. So I think it's a net positive for us from a sort of creating a pool for what we provide, and we would be the alternative that would attach the best optics along with our DSP.

Our next question comes from Christopher Rolland with Susquehanna International Group.

I guess around optical, maybe if you had a range of optical units that we might expect in 2025. Like would there be a bull case to do 2 million units or maybe even more or put another way, maybe market share? And maybe tying into this, Kishore, you talked about low power. How should we think about low power as you move to 4 nanometers, but competitors move to 3? Will you still have that advantage?

So I'll let Steve talk to you about some of the colors on the -- but I just want to add these things that, obviously, competitive advantages and power and performance are incredibly important. We take that seriously. And we are absolutely confident we'll come out with this thing, right. And that's the secret sauce of our design and architectural capabilities. That's number one. And the other part, this whole thing is like last I checked, the latest report, the 20 million units of transceiver modules that shipped and we have told you that 1 million units we have shipped. So I know it's not about 10% right now. I'm pretty proud of the 5% market share, right? That's the way I would give it. But beyond that, we don't provide color on this because there are various uncertainties on timing and that sort of thing. I would stand by the guidance that Steve earlier talked about $60 million to $70 million, yes, you should expect that. Can we do better? I desperately want us to do much better than that. So that much I promise you.

Great. And Kishore, maybe just revisiting Tore's question and all this concern about DeepSeek this week. I mean it had your stock off quite a bit itself, maybe not as much as others, but was still off. And I understand your comments about democratizing AI, but it seems like democratizing AI on open-source hardware is not very networking intensive. And so I just kind of wanted to revisit this, how these more efficient architectures might affect either inference in your opinion or training in your opinion, and mega clusters, for example, that seem to be very optically intensive. Just in terms of DSPs, units, and transceiver volumes for you guys, and if there was some sort of a reset in order rates, when would we know? It doesn't sound like you've seen anything over the past week in terms of a reset. But any thoughts on how this will play out or when we would know?

Well, this week is too short, number one. Number 2 is that, look, these dynamics are beyond my understanding. And all I know is that we make extremely good products, very low power, high performance. And the demand is huge enough even otherwise before the AI world happened, that's when we started on this road map, right? There was ChatGPT before, right? Nobody knew there was ChatGPT and that would drive the markets. Likewise, I don't know what DeepSeek is going to do, but all I know is that even if there were no AI networks present, the market was very, very huge. So from my point of view, it's a fantastic market. There is no reaction from me in any direction, stay the course. The market will be exciting for us, for MaxLinear. And I really don't have thoughts about this. But I don't think we should be surprised by disruptive innovations, and that will make us more competitive. And we have a huge appetite as human beings, right? We just eat whatever comes how much how cheap it is, right? So we'll just gobble more of it, but the TAM dollars, I don't expect to reduce, okay.

Our next question comes from the line of Tore Svanberg with Stifel.

Yes. I just had a follow-up, Kishore, because there's a lot of talk about your optical DSP business, but you've also announced getting into AEC, and ACC. There's obviously a discrete TIA market out there. So can you just talk a little bit about that? When we think about that $60 million, $70 million, is that predominantly optical DSP? Or are you also starting to see some contributions from AEC, ACC, and discrete TIAs? And one of your largest competitors just announced MPO here before the end of the year. So I assume, given your capabilities, you are now probably working on MPO as well, right?

Absolutely. I mean MPO is just a derivative of what we do as a larger DSP. So I don't know why people make such a big deal about MPO. Anybody can do this who has got a DSP PAM4. That just means there are only 3 people, but that's still, right, number one. Our revenues, predominantly, are 800G and some 400G, and there will be some AEC. Having said that, the AEC market is still pretty small. And we are -- the market being small and one data center until now trying to deploy AECs, it's still -- the verdict remains unclear whether it's going to be sort of across-the-board promulgation because there are dynamics with the copper side that are quite different. So without getting into the details, our own revenues are dominant going to be 800G, 400G, and some AECs. We do have AECs that have already qualified, and so therefore, we expect revenues to start. How big? I don't think it's still a big enough market where that would overwhelm any of the optical revenues that we will be generating. So regarding the MPOs, I think I've answered that question. ACC is even more tiny. And I think that verdict remains very questionable. AECs are questionable in terms of market size, ACCs propagation. But we have our eyes set on all of these targets, right? And we're not pulling back. We are doing everything design-wins and so on and so forth. But I'm just being an honest assessment of our own revenues where they are. So the answer to your question is affirmative. On the TIAs, right, clearly, we are one of the 3 DSP vendors on the optical side. And we work with partners on the TIAs, but it is foolhardy for us to think that we will be able to sell TIAs to our other 2 competitors on their platforms, right? So the expectation for our TIAs is so much more predominantly controlling our own platform and destiny and being cost competitive and power competitive, not as much as trying to build a TIA business. So I'm just making sure you understand that TIAs as a revenue stream are really attached to our DSPs and not -- I'm not aware of anything outside of that environment in terms of our go-to-market plan.

There are no further questions at this time. I'd like to pass the floor back over to Kishore for closing remarks.

So thank you, operator, and I want to once again thank every one of you. Hey, it's Happy New Year. You all sounded muted. Just wake up. It's all exciting times moving forward. And I wish you a happy New Year once again. And this quarter, we'll be presenting at a number of financial conferences and virtual events. We'll post the details on our Investor Relations page. Thank you very much, and Happy New Year once again to all of you.

This concludes today's teleconference. You may disconnect your lines at this time. Thank you for your participation.