Q3 2023 Maxlinear Inc Earnings Call
Participants
Kishore Seendripu; Co-Founder, Chairman, CEO & President; MaxLinear, Inc.
Leslie Green
Steven G. Litchfield; CFO & Chief Corporate Strategy Officer; MaxLinear, Inc.
Ananda Prosad Baruah; MD; Loop Capital Markets LLC, Research Division
Christopher Adam Jackson Rolland; Senior Analyst; Susquehanna Financial Group, LLLP, Research Division
David Neil Williams; Senior Equity Analyst; The Benchmark Company, LLC, Research Division
Gary Wade Mobley; Senior Analyst; Wells Fargo Securities, LLC, Research Division
Karl Ackerman; Research Analyst; BNP Paribas Exane, Research Division
Nathaniel Quinn Bolton; Senior Analyst; Needham & Company, LLC, Research Division
Ross Clark Seymore; MD; Deutsche Bank AG, Research Division
Suji Desilva; MD & Senior Research Analyst; ROTH MKM Partners, LLC, Research Division
Tore Egil Svanberg; MD; Stifel, Nicolaus & Company, Incorporated, Research Division
Presentation
Operator
Greetings. Welcome to MaxLinear's Third Quarter 2023 Earnings Conference Call. (Operator Instructions) Please note this conference is being recorded. I will now turn the conference over to Leslie Green, Investor Relations. Thank you. You may begin.
Leslie Green
Thank you, Sherry. Good afternoon, everyone, and thank you for joining us today on today's conference call, to discuss MaxLinear's third quarter 2023 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 questions. Our comments today include forward-looking statements within the meaning of applicable securities laws, including statements relating to our guidance for the fourth quarter 2023, including revenue, GAAP and non-GAAP gross profit margin, GAAP and non-GAAP operating expenses, GAAP and non-GAAP tax rate, GAAP and non-GAAP interest and other expenses, 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 product and geographic markets, including, without limitation, statements concerning opportunities arising from our broadband, infrastructure, connectivity, industrial multi-market, as well as inventory levels, the timing for the launch of our products and timing of opportunities for improved revenue and market share across our target markets.
These forward-looking statements involve substantial risks and uncertainties, including risks outlined in our Risk Factors section of our recent SEC filings, including from our Form 10-Q for the quarter ended September 30, 2023, which we file today. Any forward-looking statements are made as of today, and MaxLinear has no obligation to update or revise any forward-looking statements. The third quarter 2023 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, interest and other expense on both a 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 a reconciliation of non-GAAP guidance for future periods because of the inherent uncertainty associated with our ability to project certain [future] charges, 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 to be useful to investors as it reflects how management measures our business.
Lastly, this call is also being webcast, and a 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 Seendripu
Thank you, Leslie, and good afternoon, everyone. In Q3, our revenues were $135.5 million. And non-GAAP gross margin was 60.8%. Infrastructure revenues, specifically wireless infrastructure was the main highlight, up 1% sequentially and 40% year-over-year.
Our financial results and outlook continue to reflect the channel inventory overhang, especially in the broadband and WiFi markets. We expect the FX of the inventory to persist into 2024 and Steve will talk more about the actions we are taking to align our financial structure. We continue to make strong progress growing our infrastructure business.
Infrastructure represents $50 million in Q3 revenue and has grown substantially since its inception only a few years ago. With the infrastructure in wireless, the expanding 5G global rollout of new millimeter wave backhaul technologies and multi-band and hybrid millimeter wave and microwave backhaul radios, is allowing us to significantly increase the silicon content per platform of our modem and RF transceiver products.
In our high-speed optical data center interconnect market, the ongoing adoption of AI in the cloud is driving exciting design win activity for our 5-nanometer CMOS, Keystone 800 gigabit optical PAM4 solution. We have ongoing qualifications in multiple hyperscale and enterprise opportunities for which we expect to begin ramping in mid-2024.
During Q3, we also announced a new member of our Keystone family, a 5-nanometer Keystone PAM4 DSP per 400 gigabit and 800 gigabit application, with integrated VCSEL laser drivers. This product enables best-in-class power consumption for 800 gigabit short-reach optical transceivers and active optical cables for data centers, AI and machine learning platforms. And high-performance computing applications.
Earlier this month, at the OCP Global Summit in San Jose, we also demonstrated our Keystone solution supporting active electrical cables or AECs. Keystone is the only 5-nanometer product with DSP available today for the AEC market, providing best-in-class power consumption and programmability.
We're also making exciting progress with our Panther III series of storage accelerators for the enterprise all-flash array and hybrid storage appliance systems. We've entered initial mass production ramp with the (inaudible) leading enterprise storage appliance maker and have visibility into additional new design win volume production ramps next year. We expect this business to double in 2024, with continued strong growth in '25 and beyond.
Turning to broadband, despite the near-term challenging environment, the longer-term outlook for the broadband access networks is solid as the industry migrates from Legacy DSL and older PON technologies to 10-gigabit PON fiber access. We continue to ramp with the major North American service provider and are layering additional design wins, including another Tier 1 service provider, which will begin to contribute revenue in 2024.
With our industry-leading single-chip integrated fiber PON and 10-gigabit processing gateway and connectivity solutions. We expect continued strong design win traction leading to a multiyear fiber broadband growth cycle. Recently, we also announced Puma 8, our DOCSIS 4.0 system-on-chip cable modem and gateway platform, which enables the speed, latency and low power consumption necessary for next-generation 10-gigabit service rates for MSOs. We expect to see initial DOCSIS 4.0 launches in the market as early as the end of 2024.
In connectivity, the design activity for our Wi-Fi 7 is robust, and we anticipate early revenues coming in the second half of '24. Wi-Fi 7 has the enhanced ability to efficiently manage the increasing number of connected devices, which have grown tenfold since 2018 and the higher bandwidth requirements in the home. As a result, globally, service providers are embracing the transition to Wi-Fi 7 to improve both user experience and performance. For MaxLinear, Wi-Fi 7 has the exciting potential to drive significant ASP growth and higher attach rates in our Broadband Access platforms versus previous generations.
Moving to Ethernet connectivity. We continue to build on our core portfolio of 1-gigabit Ethernet and 2.5-gigabit Ethernet PHY technologies. We not only offer single and quad 1-gigabit Ethernet and 2.5-gigabit Ethernet PHYs. But in Q3, we started sampling the industry's first Octal 2.5-gigabit Ethernet PHY-Switch product. This new product family of switch, significantly expands our addressable market by $300 million through 2027 and addresses both enterprise and SMB switch markets. As well as the gateway and router markets.
We expect today's more than 2 billion copper, 1-gigabit Ethernet ports in the market or existing CAT5 cabling to transition to an optimized and enhanced 25-gigabit Ethernet offering. With a strong design win activity, we expect to begin revenue ramp in the first half of 2024. As we head into 2024, we are laying the critical groundwork for future growth with robust design win activity and continued technology innovation.
Even as we drive operating efficiencies to navigate near-term headwinds, we are excited that many of our investments, over the past several years in infrastructure, broadband access, connectivity in high-performance analog markets are now poised to contribute meaningful revenue.
With that, let me turn the call over to Steve Litchfield, our Chief Financial Officer and Chief Corporate Strategy Officer. Steve?
Steven G. Litchfield
Thanks, Kishore. Total revenue for the third quarter was $135.5 million, down 26% versus Q2 and down 53% year-over-year. Broadband revenue was $34 million, down 36% versus Q2 and down 71% year-over-year. Connectivity revenue in the quarter was $15 million, down 60% sequentially and down 82% year-over-year.
Our infrastructure end market continued to grow in Q3 as a result of solid demand and growing market opportunity. Infrastructure had revenue of $50 million, up 1% versus the prior quarter and 40% year-over-year.
Lastly, our industrial and multimarket revenue was $36 million in Q3, down 16% sequentially and 24% year-over-year. GAAP and non-GAAP gross margin for the third quarter were approximately 54.6% and 60.8% of revenue. The delta between GAAP and non-GAAP gross margin in the third quarter was primarily driven by $8.3 million of acquisition-related intangible asset amortization.
Third quarter GAAP operating expenses were $91.8 million, including stock-based compensation and performance-based equity accruals of $8.2 million, combined and acquisition and integration costs of $2.2 million. Non-GAAP operating expenses in Q3 were $75.1 million, down $7.3 million versus Q2 at the low end of our guidance range. Non-GAAP operating margin for Q3 was 5%. GAAP interest and other expenses during the quarter was $23.7 million, driven by the ticking fee from the debt commitment associated with the terminated Silicon Motion transaction.
Non-GAAP interest and other expense during the quarter was $5.3 million. In Q3, cash flow using -- used in operating activities was $12.8 million. We exited Q3 of '23 with approximately $203 million in cash, cash equivalents and short-term investments. Our days sales outstanding for the third quarter was approximately 106 days, up from the previous quarter due to shipment linearity.
Our gross inventory turns were 1.4%, down from Q2 levels. As Kishore mentioned, we've taken meaningful actions to align our cost structure with the current environment, and we expect to begin to see the benefit in Q4. These actions include a head count reduction, site consolidation to drive efficiency and scale at our primary sites and more prioritization around the projects, that we believe will drive growth over the coming years.
MaxLinear has a solid track record of managing our business through downturns with strong fiscal discipline and focused spending. This concludes the discussion of our Q3 financial results. With that, let me turn to the guidance for Q4 of 2023. We currently expect revenue in the fourth quarter of 2023 to be between $115 million and $135 million.
Looking at Q4 by end market, we expect connectivity and industrial multimarket to be up and broadband and infrastructure to be down quarter-over-quarter. We expect fourth quarter GAAP profit margin to be approximately 52.5% to 55.5% and non-GAAP gross profit margin to be in the range of 59.5% and 62.5% of revenue. Gross margin continues to be stable despite lower unit volumes with the range being driven by the combination of near-term product, customer and end market mix.
We expect Q4 2023 GAAP operating expenses to be in the range of $125 million to $135 million. We expect Q4 2023 non-GAAP operating expenses to be in the range of $72 million to $78 million. We expect our Q4 GAAP and non-GAAP interest and other expense to be negligible. We expect our Q4 GAAP and non-GAAP diluted share count to be between $82.5 million to $83.5 million.
In closing, we continue to navigate a dynamic environment in Q4, but are laying important groundwork in strategic applications that will drive our future growth. Our solid product innovation and execution in Wi-Fi, fiber broadband, access gateways, Ethernet and wireless infrastructure has positioned us well across a number of exciting emerging markets. As always, we will continue our strong focus on operational efficiency, fiscal discipline and shareholder value, as we optimize for today and plan for an exciting future.
With that, I'd like to open up the call for questions. Operator?
Question and Answer Session
Operator
(Operator Instructions) Our first question is from Tore Svanberg with Stifel.
Tore Egil Svanberg
My first question is on broadband and connectivity. Obviously, these are bulk down quite materially year-over-year from this inventory digestion. It does sound like connectivity may have found the bottom as you're guiding for growth there. First of all, I just want to make sure you confirm that. And then second of all, on broadband, are you seeing any sort of bottoming at all in that business given its $34 million run rate?
Steven G. Litchfield
Yes, Tore. So you're right. And I think you're also rightly looking at broadband and connectivity. Somewhat together, they have been influenced by some of the similar dynamics of just the massive amount of inventory that we've seen in the channel. It's definitely persisted more than, I think, what we had anticipated. But you're correct in -- we did say that we would see connectivity start to move up a little bit in Q4.
I think for both broadband and connectivity, while connectivity has some growth drivers, as we've talked about over time, that are a little bit better. I mean both are still being influenced by inventory in the channel, and I would expect to see that last through kind of Q1, maybe even some residuals in Q2, but we are seeing an improvement. I mean, we talked last quarter about bookings. I think we're starting to see some bookings. It's still early days. They're not super robust by any means, but there are some encouraging signs, nonetheless.
Tore Egil Svanberg
Great. And on infrastructure, obviously, that was the highlight this quarter. You're expecting that to come down in Q4. I'm just hoping you could talk a little bit about the extent of the decline. And is this basically just sort of like volatility because obviously, it was up a lot and maybe it's coming back down. But yes, in any way we should think about infrastructure for the next few quarters would be great.
Steven G. Litchfield
Yes. So infrastructure, I mean, I think we've been really pleased with infrastructure. I think we're making a tremendous amount of progress in a business that we've been investing in heavily for a long time. I think as we spoke over the last couple of quarters, the first half of the year was extremely strong. That carried through into Q3. But we still see some of those product ramps, particularly on wireless infrastructure, kind of slowdown that can be lumpy. And so we can see some slowness in Q4 that probably carries over to the beginning of the year before it starts to pick back up and grow in 2024.
Kishore Seendripu
And there are several contributors beyond wireless, right? I mean, in fact, optical data center investments, we're very confident in the progress we're making. We expect to see ramps in the middle of next year. We're going through interop cycles right now and a lot of design win activity going on, whether it is in transceiver products or in active optical cables or even active electrical cables, right?
And then over on top of that, there is our storage accelerators, which we have not spent time much, a lot of great design win traction, a very, very strong design win pipeline. It's going to double next year into high teens, if you will. And then beyond that, is going to keep growing very robustly for the next few years.
You want to keep in mind that infrastructure lasts -- revenues last a long period of time. That's why they're -- so likewise, they have a long leading time. So on the -- all in all, our anticipation is over the next 5 years, we should be able to get to double our infrastructure revenue from today's $200 million run rate. To maybe in the $0.5 billion vicinity. I mean, that's the ultimate goal we are after, and we feel that we're making very good progress towards it.
Operator
Our next question is from Quinn Bolten with Needham & Company.
Nathaniel Quinn Bolton
I guess maybe a follow-up on Tore's question. Obviously, this inventory correction has lasted longer and causing a deeper revenue trough. But as you look to next year, can you give us any sense how much of your ownership in the channel -- these businesses -- connectivity and broadband down well more than 50% peak to trough.
Do you have any sense what natural run rate demand is? What kind of snapback might you see, once we flushed this current inventory? And then just any thoughts on all of the government spending or government funds that are available for broadband infrastructure? When does that start to benefit this business?
Steven G. Litchfield
Yes, Quinn. So you're right. I mean, it's definitely been worse than anticipated. I mean there's been a big build-up. We've seen some kind of bad practices, some overbuilding definitely over the last couple of years. That's kind of playing out right now. The common question that we get is what is it recover to and when? I guess I would just say that we're confident that we're kind of seeing the bottoming here.
I -- you raised a couple of good points about the government spending that we've highlighted. We have seen CapEx commitments. You've seen a great transition to PON on more rural areas, deploying more broadband and more broadband upgrades. I think we continue to see that the outlook that our customers and the service providers in general, have is the outlook does remain very good.
So thus, our excitement in some of Kishore's remarks around WiFi, Wi-Fi 7, some of our PON business. So there are some exciting things going on. some of the bigger money like the bead money and some of that just part of that Infrastructure Bill starts to be deployed, I should say, it gets allocated at the end of next year.
So it's still a little ways off, but we're seeing ramped upgrades in Europe, in the U.S. over the next 2 to 3 years. So we still have good visibility on this and do expect to see these upgrades. But in the meantime, unfortunately, we're having to work through this inventory.
Nathaniel Quinn Bolton
And you guys thought last quarter that September would be the bottom. Obviously, you've updated that guidance today with December being down. Are you willing at all to make a comment, do you think December is the bottom? Steve, you mentioned inventory overhang probably continues through at least Q1 and maybe residually into Q2. Could -- could Q1 be down from the fourth quarter? Or is it just too early to call? And then I'll have a quick follow-up.
Steven G. Litchfield
Yes. So if I recall correctly. So last quarter, we talked about Q4, expecting somewhat of a modest improvement, but we didn't say we would be out of the woods on the inventory side. We expected inventory to last into next year. But that being said, we thought we would see more of a recovery, and we didn't see that. So that's correct.
As we look out into next year, what's the shape of that look like? It feels like we get back to the normal seasonality in our business. I wouldn't be surprised to see softness in Q1, but then you start to build off of that as we normally would. Our Q2 and Q3 -- Q3, as you know, have been historically much stronger than, say, Q1 and Q4.
And so I think there are some dynamics of seasonality that are starting to play a role again. But I think more than anything right now, it's just getting through the rest of the inventory that's sitting in the channel.
Nathaniel Quinn Bolton
Got it. And then just quickly, you mentioned some of the actions you're taking, including reduction in 4 site consolidation, prioritization of certain projects yet. Your guidance for OpEx, I think, is flattish sequentially in December, kind of in line with, where we were already looking for OpEx to be in Q4. And so I guess I -- I didn't see much of a change in the OpEx outlook. Are these options -- sorry, the cost reduction plans you've talked about, does that kick in really more in the first half of next year is $75 million the right run rate kind of as a baseline? Or do you think it could move lower next year?
Steven G. Litchfield
So, I'll comment a little bit on just overall expense reduction efforts, right? So we did start early in the year. We've made some changes. And then more recently, we've made some additional changes that will start to impact Q4 and beyond. So you don't see so much in the Q4 time frame just because it's beginning now.
You'll see that continue to come down throughout next year, pretty linearly. Sometimes that's masked a little bit by some of the NRE that we take as an offset to OpEx. And -- but I would expect to see the overall OpEx number decline throughout the year. And so I think -- I mean, if I was to put a number on it, it's probably $285 million to $290 million next year. So that's the kind of the size of the decline, and that also offset some additional NREs that we had taken, which were much larger in 2023.
Operator
Our next question is from Gary Mobley with Wells Fargo.
Gary Wade Mobley
Looking at your largest customer, it looks like they purchased about $90 million of product from you in the third quarter of last year. That's probably down to something less than $10 million in this most recent quarter. So my question is, what's the right level for purchases with this customer, which I presume is representative of your broadband cable and WiFi business.
And does your current fourth quarter guidance and maybe your longer-term view, contemplate the transition of this business away from that customer into the hands of somebody else?
Steven G. Litchfield
So Gary, I mean, with regard to the -- some of the disclosures and the top customers, we've had a very large customer for a while. I think you've heard us talk about our top customers are often kind of more -- I'm going to refer to them more as kind of box vendors that kind of sit in between us and the operator. And I think you're well aware of most of these service providers are looking to kind of keep multiple silicon vendors in the mix so that they've got some leverage.
We don't see that changing much. And so we don't necessarily put too much weight on these customers, they ebb and flow from time to time. And even this year, we've actually seen some of our top customers switch as different service providers start to ramp. We've actually seen that through this year, and we've seen some nice improvements on some of these newer customers.
Gary Wade Mobley
Okay. Question for Kishore, as my follow-up. Kishore, you mentioned in your prepared remarks, some design win traction, I think, is how you phrased it for Keystone in the data center market. So I think you also mentioned a revenue ramp into the second half of the year. How much visibility do you have at this time in that revenue ramp? And how influential can some of these early days design wins be for MaxLinear?
Kishore Seendripu
So I think there are 2 parts to this question. I mean the visibility is very, very clear, right, in the sense that you work directly with the OEMs or module makers, where these transceivers are cable manufacturers, optical cable and you directly work with them. And you work with them because they're being aligned up to utilize our silicon at the endpoint, which is usually the data center folks. And so you have direct visibility now.
The timing of when each one ramps and how the distribution and portioning of the revenue goes is the one that you have a little bit of what I call uncertainty of. However, the fact that we are in the mix, the fact that [certain calls] are going well. And the interactions, the commitment to take us through all the qualification process, the direct visibility you have.
So having said that, like I said in my prepared remarks, the interops and [calls] are going on still, and we feel very good that our silicon is sound and strong and interops will go favorably. At this stage, that's our conviction, and that's our, what I call reading the tea leaves, if you will.
Regarding how much influence they will have on our revenues, absolutely right. Even if you were to -- you don't do a top-down game plan, you always have to do a bottom-up game plan for revenues. And -- but if I were to map all of that, we hope to expect about 20% share sometime in the 3- to 5-year window of each of our customers. That's our base plan. And if you exceed that, you'll do much better.
So how big can the business be in 5 years? Obviously, it can be somewhere between $100 million to a few hundred million dollars, right? That's a wild card here. So yes -- and that's the basis on which including wireless, optical, the accelerator business is where the confidence comes, that in a 5-year window or infrastructure business should be in the ballpark of the $500 million range, right? And that's the goal. And all of which, I would say, optical and storage activities of the greatest growth curve ahead of them.
Operator
Our next question is from Christopher Rolland with Susquehanna.
Christopher Adam Jackson Rolland
I guess the first one is just kind of the swings that we've seen here from peak to now trough, going from $105 million in connectivity to $15 million, for example, have just kind of been incredible. So I guess, first of all, do you guys really view this as all inventory digestion? Like are we done in connectivity? I don't see how it can get too much kind of lower than this.
But do you have any idea of how much extra inventory is out there in the channel? And then moving forward, are you guys rethinking kind of systems to judge this inventory level that's out there? Are there new kind of processes that can be put in place to have a better view?
Kishore Seendripu
So if we were not analyzing the channel and the inventory, we shouldn't be in this job, right, in the first place. So obviously, we are analyzing these 2 debt. And sometimes it's very difficult because there is a steady level of guarded -- disposition from your customer to their customers. So the closer they are to you, the more information you get a little bit more accurate.
But I just want to hark back a little bit more. One of the most important things we need to start guessing are making educated guess is about when did the over build start? It started, let's assume the over shipments happened in the pandemic period over the last 2 to 3 years. And if you think that there was a, let's assume a 30% over shipment then you're looking at probably 3 quarters worth of at least actual and demand that is sufficiently provisioned? And how far are we into it? Maybe we are into a core profit. right?
If that is the logic you go run through as all logical people should, then you should start expecting a recovery somewhere in the second half of next year. Can you dial it in within a quarter? No. The second part of it is like, have the dynamics in the business change? No. I mean, always, we get excited about the greatest and the greatest new offer in technology, blah, blah, blah, and our customers talk about it and our customers' customers talk about it because I hate to say that's what the investors want to hear. Okay. But the real revenues are generated by all the products. Products that are actually long term, they're sticky because of software or performance or whatsoever.
And they're also cost down for the customer so they can ship more of it. In that sense, the dynamic isn't in the marketplace in terms of we have a robust Wi-Fi 6 portfolio. We have a robust Wi-Fi 7 offering. We have the SoCs to complement our broadband Wifi offering, right? I know that the broadband access, it's funny, when it does well, it's get discounted, when it goes down -- well, that's the problem. Sort of think -- no, honestly, if you step back, we're trying to build a broad-based portfolio company with potential to generate large profits and earnings per share for our shareholders. And at the same time, build scale and while investing in what I call more resilient businesses like our infrastructure and so on and so forth, so that we can build a comprehensively a large company. I mean, that's the ultimate goal. And no matter what happens now, we'd really be focused on the long-term goal, and that we're very committed to.
Christopher Adam Jackson Rolland
And then secondly, about your infrastructure business. The upside there, you guys kind of highlighted millimeter wave and 5G backhaul. I would say, first of all, on the millimeter wave side, I think it's been very slow adoption. So it's interesting to see, you guys picking up why now? And then secondly, on the 5G side, we've seen builds really start to slow even in India now. What are the specific programs that you guys are linked to, that are kind of swimming upstream here?
Steven G. Litchfield
Yes. Just real quick, maybe a clarification, Chris. I think you're aware, most of our business has been backhaul. So these backhaul transceivers that we've been shipping this year has been a big driver, right? So very different than the markets, the access -- 5G access markets that you're describing. And so these are microwave backhaul that's replacing fiber kind of in between base stations. So very -- again, I just want to emphasize a very different market than the decline that you're referencing, which we also see. We sell into the wireless access market. We are familiar with it, not surprising. There's definitely been weakness there. So -- but we've been able to gain traction in some of these other markets. They're a little [nicheier] markets. They're a little bit smaller, but they've been great growth drivers for MaxLinear.
Kishore Seendripu
There is -- there's a trend in the microwave, millimeter wave backhaul deployments as well, right? People are deploying more and more multi-band deployments and multi -- multi-band I mean, like millimeter wave band and microwave kind of combined radios or hybrid radios. So that increases the content as well.
Yes, India has been a driver as India was rolling out strongly in 5G. And now we see a slowdown. So we expected the slowdown, as I've guided accordingly. And you also have to keep in mind that, we did not have excess inventory in the infrastructure channel. We basically -- we're running short on supply, and we supply to the market.
So the growth you're seeing, what you call our ability to pull a [salmon] go upstream is really based on the fact that the channel was not over stocked. So we are shipping to natural demand. Now with the slowdown, we will be caught up with the slowdown as well.
So -- so really, the growth is coming in the backhaul to these multi-band hybrid deployments, which millimeters a part of. And you also want to think about the fact that as the access bandwidths increase, the fronthaul and backhaul data pipe is no longer going to be provisioned sufficiently by microwave and they have to use millimeter wave where it is cost effective to -- and combined with microwave and they're making trade-off versus fiber. So you can imagine countries like India and even in the U.S. in metropolitan zones and things like that. People are trying to do a lot of hybrid deployments.
Now does it slow down? Yes, absolutely. We have guided so accordingly, and it's going to be a little, what I call the telecom CapEx being dramatically slowed down as a lot of the telecom OEMs have talked about. In fact, some are preannounced, right? We should see some impact of it, but this is where our infrastructure is going to really be driven by our growth in our storage accelerators and our optical data center investments. So I think it's turning out to be a pretty nice portfolio, which I'm quite pleased actually, those have taken a while.
Operator
Our next question is from Ross Seymore with Deutsche Bank.
Ross Clark Seymore
Just wanted to ask a couple of questions. For the fourth quarter not going up sequentially, was that demand change, more inventory was out there than you expected? And I know those 2 things are interlinked. But what changed from 3 months ago to today, that leads the fourth quarter to be down sequentially?
Steven G. Litchfield
Yes. I mean, I think it's exactly as you stated. I think it's both, right? I mean there's definitely more inventory. We saw more pushouts. I mean, we saw bookings in the quarter, so we saw some improvements. But it wasn't as much as what we had originally expected 3 months ago.
Kishore Seendripu
And honestly, we ourselves are sort of baffled, if you will, as to how much inventory is out there and with the slowdown and how to reconcile that, right? So it's getting clear as the slowing economy sort of is all the catching up with us. I think there are 2 parallel economies out there right now, a tech economy and there's a chip economy and maybe there's a consumer economy. I don't know there are maybe there are 3 of them, and we are definitely in the chip economy. And we're seeing some of the downsides of that.
Ross Clark Seymore
I guess -- second question, I have 2 quick follow-ups. The first one, for next year as a whole, I know you're not going to guide to total revenues. You talked a little bit about the linearity of it with the seasonal comment earlier. But from a high level, what do you think are the idiosyncratic tailwinds or headwinds that you guys as a company have as you look at '24?
Steven G. Litchfield
I think it's pretty straightforward. I mean, I don't -- I'm not going to guide, of course, the next year, but I mean I think the shape of it is probably the opposite of this year, right? I mean we started the year out really strong, and we saw that kind deteriorate to some degree. We're working through this inventory, and I think you're likely to see us continue to improve.
And -- and a lot of that is coming from just naturally inventory burning off, but it's also coming from new programs, new products that are going to ramp in that kind of the second half of next year. I mean, you've got a lot of new wins coming from optical. That starts to ramp next year. Wi-Fi 7 starts to ramp next year. So you've got several new programs that are going to ramp on top of the inventory just naturally burning off.
So both of those will help. I mean, I guess the only other thing that I just mentioned on another question was seasonality, I think you probably see a little bit of softness in seasonality, but I think it's more influenced at least in the short term by the inventory that sits in the channel.
Ross Clark Seymore
Got it. And my last one, and forgive me for speaking in 3. I know it's a confidential process in the arbitration with Silicon Motion. But any sort of update on either the timing, a reiteration of what you said before, but the timing of it or the potential magnitude, any sort of color on that is that tends to be the most frequent question I get. And again, I appreciate your somewhat, if not significantly limited in what you can say.
Steven G. Litchfield
Yes. I don't think anything has changed, just as we had updated before. I mean the only change is that Silicon Motion filed for arbitration, confidential process. So you're correct in that we can't add any more color there. Still expect that arbitration process to take 12 to 18 months.
Operator
Our next question is from David Williams with Benchmark Company.
David Neil Williams
Kishore, maybe you could talk a little bit more about the traction you're seeing in the Keystone platform? And what's the magnitude of that ramp do you think for next year. Is there any -- maybe just a little more color that you provide around that to help us understand the traction in growth?
Kishore Seendripu
We don't guide to the future in that sort of short time lines. But I think from where we are today, we had a pilot builds right now as the interop cycles continue. And hopefully, next year, we have something in the teens or beyond that. And hopefully, beyond that because that would be disappointing if it's in the teens.
And then -- but I could talk to you in terms of 3 years -- in a 3-year window, could we cross $100 million? Yes. I mean that would be a natural expectation, right? So can we do better than that? Absolutely, that's based on share shifts. I think 1 wild card, the timing of the deployments of multiple data centers and they transition to 100-gigabit per lambda, whether it's 400 gig or 800 gig or 1.6 terabit on the 100-gig platform, if you will.
And so you're counting on multiple players coming on. Right now, we have confirmed transitions from big -- 1 big data center guy, I know that in the [NVIDIA] AI clusters, they are deploying 100 gigabit. But you know what, it's like we have to land into it rather than win into it right now. So we are trying to win into it. And as they increase their supplier base, hopefully, we are one of the selected ones, but I'm not saying that we are selected or we are in it. So please don't mistake that. I'm just trying to say our focus is right now winning and some good luck on the share, basically, okay?
David Neil Williams
Okay. Can you say is that progressing as you would have expected? Or is it maybe a little slower than you had hoped?
Kishore Seendripu
It always been slower to me, honestly, it's been a few several years since we've been investing in the optical data center. And every time, it seems slow for me because I'm dying with anticipation, right? So -- but so far, we feel very good about where we are. And like I said, if you read the tea leaves, that should actually be more positively disposed than my forecasting will indicate to you.
Steven G. Litchfield
I think, David, another positive, I mean, you're starting to see our IP that we've developed in optical, starting to broaden out. We can go after AEC opportunities, AOC opportunities. And that really to leverage the development that we've done thus far. Those are also types of programs that can ramp quicker versus some of the other transceiver platforms. Yes.
David Neil Williams
Great color. And then maybe lastly, just on the carrier or maybe the operator side. If you look out, are you hearing anything in terms of the beginning of the year of CapEx, planning? Or is there any sense of optimism that you're beginning to hear maybe for 2024 deployments and maybe CapEx spend?
Kishore Seendripu
It's always the hardest thing in all these years in broadband. I can tell you very clearly that, they start their process sometime in Q3, and then (inaudible) know anything and Q1 -- by the time of the end of Q1, then sometimes they just go super aggressive as well. So -- but these are unique times because there's a lot of inventory sitting out there, even if they're going through that process, us feeling the impact of their OpEx decisions is going to be delayed for sure, right? Because you're going to be depleting even the inventory. So you won't feel the urgency they would come rushing towards end of Q4, or in the middle of Q1 in the past pre-pandemic period, right?
But now there's still enough inventory in the channel that it's going to be dampened. So we wouldn't be picking that signals much. But I think -- but I think we should all expect that everybody is going to be tightening their belts, right -- and so it will be subdued whatever they're going to be up to.
Operator
Our next question is from Karl Ackerman with BNP Paribas.
Karl Ackerman
Kishore and Steve. Two questions if I may. I guess I first wanted to ask, there have been many questions on this call sort of asking whether -- there are structural impairments to some of the broadband and connectivity portions of your business. So I'd like to ask specifically about your connectivity business, how much of that business today is on a rough and tough basis, split between wireless and wired? I think that would be certainly helpful as we contemplate some of the content drivers that you talked about earlier on this call as it relates to Wi-Fi 7 next year, as well as some of the growth opportunities today for Wi-Fi 6.
Kishore Seendripu
Okay. I would say we have little or no exposure to wireless access from a connectivity side. To the extent that we have exposure to the wireless access on the broadband connectivity side for our WiFi's offering. It's really in the telco platforms where they tend to be the gateway box, where, for example, they have a fiber PON chip with our gateway processor and a WiFi and they will be what I call an input to that box that comes from a 5G access (inaudible).
So very little or no exposure to wireless broadband access, gateway access, okay? So if you take out that on the wireline, we have pretty much 100% of our exposure to wireline access. However, it may be the, let's say, 90% of it and 10% of it is what we call stand-alone router gateways that we are -- we started making progress at the beginning of the year to get revenues, that are outside of the operator gateways, right? That would be the landscape. So you should -- on a practical terms, you should associate 90% of our WiFi connectivity reviews with our wireline broadband access gateways okay?
Steven G. Litchfield
And maybe, Karl, just to add on a little bit. So don't forget on connectivity side, we also have about Ethernet. So while a lot of our declines in the gateway, I mean, it's been driven by both WiFi and Ethernet. One of the big things going on is we're seeing more exposure, Kishore shared some commentary around what's going on with this transition to 2.5 gig. We are seeing a lot more interest in that, and we do expect to see more growth in '24 and '25 from Ethernet. As well as, of course, the WiFi opportunities.
Karl Ackerman
That's very helpful. I appreciate that color. For my follow-up question, I guess, are lead times in backlog back to normal, I guess, pre-COVID levels. Is that a fair -- fair way to think about it today? And then second, it's nice to see the decline in inventory. But any thoughts in terms of a target level of inventory, as you look to manage expenses over the next couple of quarters?
Steven G. Litchfield
Yes. It's a good question. So I think we've been doing a really good job on inventory, as far as bringing it down. but we got to do better. And we will continue to do better. We jumped on this pretty quick going all the way back to the tail end of last year.
But unfortunately, the revenue decline has just been such that we've not been able to get the inventory out as fast as we'd like to. With regard to lead times -- with regard to lead times, I would say we're kind of back to normal. I mean we're kind of quoting 16-, 18-week lead times, which is pretty typical in our business. I mean there's couple of businesses that are probably a little faster than that, but that's kind of our normal. So I wouldn't say that, that is problematic, whatsoever at this point.
The backlog also -- backlog and bookings. So bookings, as I've talked about, I mean, have been very low because you had super good backlog for well over, I don't know, almost 2 years now. And so now I feel like we're getting through that adjustment phase, where you're going into a quarter with whatever, 50%, 60% backlog, whereas you -- historically, you've been running for the last 2 years, you've been running at 100% backlog, and that's changing. And that -- that getting back into that normal rhythm, that's what we're used to. That's what we know. And so actually, looking forward to kind of getting back. Some of the uncertainties has been around backlog pushing out of a quarter. That's really where the problem has been.
And so we're starting to see that improve.
Kishore Seendripu
But to be honest, Steve, if to the extent I can say we want 0 inventory in our books so that we get the PO bookings going on our customer side, I bet you I will support you on that.
Operator
Our next question is from Ananda Baruah with Loop Capital Markets.
Ananda Prosad Baruah
Maybe just kind of dovetailing right off of that last topic with Karl, it seems like everything you guys just talked about would suggest that new normalized inventory levels at customers, exiting this would be the same, as they previously were going in. Do you think that that's based on what you can see a fair assumption?
Steven G. Litchfield
I'm not sure that I followed you, Ananda. I'm sorry.
Ananda Prosad Baruah
Well, your customers hold, I mean they probably view their inventory level -- to them, they have what they would interpret as normalized inventory levels, I would imagine. And that was -- that looked a particular way going into 2020, they can change what that looks like, Steve, but it sounds like -- I mean, they can change it higher, they could change it lower, they could be with the same coming out of this. But it sounds like based on what you're saying. Well, here's my thought. So does that make sense, though? What I'm saying like they probably say, we want to hold some number of weeks of MaxLinear inventory?
Steven G. Litchfield
Sure.
Ananda Prosad Baruah
Yes, yes. So it sounds like it sounds like what you're saying is the lead times are back to pretty typical, and you're at 50%, 60% backlog, it sounds like ship out is meaningfully higher than ship in because you have typical lead times. So it sounds like they're operating you guys and typical lead time is working down the inventory, to the 50%, 60% backlog. So that would sort of suggest to me -- I guess, it really kind of crux of the question. It sounds to me like they're already working you guys as if getting you back to pre-COVID inventory levels. I was just wondering if you have an opinion on where that might settle in, that would also inform when you guys inflect.
Steven G. Litchfield
Yes. Look, if I understand your question correctly, I agree that the -- I mean I think the whole industry sees that lead times have come down. And so they're waiting. They're taking more risk. They actually kind of -- not sure if we're seeing eye-to-eye on this, but I think customers do have, still have -- if you look across the industry, there's still a lot of inventory either in the channel or even sitting at end customers. And so while that is still high, I mean, they are still burning that down. And -- but we're getting back to those normal times, as I stated. I mean I think it's another couple of 3 quarters, but we are seeing improvements.
Ananda Prosad Baruah
Yes. And I guess the question really was, do you think that they settle you in at lower levels than previously, when things get back to normal, do you think that...
Steven G. Litchfield
I think -- yes, I mean, absolutely. That's what we see in every cycle. It swings the other way, right? They will take more risk. They will wait too long. That's exactly what's going on right now and in my opinion across the entire industry is that they are waiting because they -- either they know or they think they know that there's enough inventory out there. And so they're going to wait as long as they can and nobody's -- every one of these customers, their operations guys getting pounded on for having too much inventory. So they're going to not order and they're going to risk being late. And in this environment where demand is kind of so that's probably okay.
Ananda Prosad Baruah
And then just a quick follow-up. Kishore, you mentioned a couple of times about the storage accelerators and potentially picking up in '24 and then potentially to be robust for a few years. Do you -- in any context around how impactful those could be?
Kishore Seendripu
It's very impactful, right? I mean generally, by and large, whenever we invested in any new product areas, we hope to build at its peak run cycles somewhere in the $50 million to $100 million per year product, if it's an infrastructure product out, otherwise, the economics don't make sense.
And there's a rhythm to it, what is the next product going to be built, and that's sort of the thing that's sustaining revenue and growth, right? -- self-sustaining growth and revenue. So that is the expectation for this product that it's going to be somewhere in the $50 million to $100 million per year revenue when it hit its peak revenue.
So eventually, you have really rapid growth, like we said, double next year and maybe grow 50%, 60% the following year from there and maybe it hits the peak point in some time in the third and the fifth year from now. right, hopefully, a third year, not the fifth year, and then it holds there for a long time, and we launched new products.
And the more important thing, and this is where -- this is the key point here, and maybe we never ever connected the dots very well here is that, there is a place and need for these accelerators in the cloud market as well. And as the AI and the cloud and the edge increase accelerators will become essential to it. So right now, we're investigating partnerships with AI vendors, if you will, where there could be a joint offering, and we're seeing a lot of what I call openness for that joint collaboration of -- join this thing.
And that really is the key to the storage business is the enterprise is going to be a massive consumer of storage. And latencies and access speeds and all of those are going to be incredibly important at the [ads] and even inside the cloud, moving forward, even with the AI is going to get even worse. And really, this is a play that goes together with all the offerings on the AI processors as well. So that's the next step in the evolution of our storage accelerator business. And maybe this is the first time I really connected the dots in that sense for you.
Operator
Our next question is from Suji Desilva with ROTH MKM.
Suji Desilva
Question about the carrier PON program. I'm curious if that is -- the rollouts happening as you had expected or whether that carrier is pushing out or delaying that rollout because of the macro environment?
Kishore Seendripu
The role is already happening. We've been shipping for this whole year actually. The only difference is that the earlier part of the year, the rollout was slower than we expected. So they were sitting on a bunch of inventory, and then they started shipping. Now they're shipping on a natural cadence. And we are already working on the next-generation platform, but my expectation is the next generation is going to be delayed and the existing generation platform with 10-gigabit PON XGS-PON and Wi-Fi 6 or 2, which is the enhanced throughput 1 is going to have a long life. Having said that, we've got the next-generation offering as well.
So I don't think there's any change in plans yet. There will be natural slowdowns and seasonalities and that sort of thing, but nothing out of the ordinary because to start with, there is not as much inventory on our side as was in the beginning of the year.
Suji Desilva
Okay. That's helpful, Kishore. And then lastly, on the AEC market within optical, that's kind of coming online here. Can you help us just think about the relative size? Do you think that market will be a year or 2 out versus the AOCs and the passive cables. Just to understand how big you think that market grows as a share of the cabling?
Kishore Seendripu
That's a very, very interesting question. I know there's a lot of excitement in AEC, but AEC has been a very, very tiny market looking backwards -- right? The whole rationale for AEC comes in as the speeds have increased dramatically, where passive copper cannot meet the performance, and you need active electrical cables.
There's a place for AOCs as well. If I'm in an AI cluster, I don't want to do anything with AECs, right? I want to go optical. So it's really the mix and match. So this market can be huge, gigantic. And I think any forecast will underestimate the volume it can be over the long term. It's just like USB 3.0, right? It can just keep growing.
However, the prices are going to be a lot more optimistic than they really will play out to be, right? All in all, I expect the market, I don't know, anywhere between $200 million size for the chip guy to as big as $1 billion for transceivers, active optical cable and active electrical cable combined.
But there is -- these cards in the last 20 years of existing is MaxLinear for me. I've never seen a single chip supply in the Communications segment more than $300 million of product. So the $1 billion TAM, $300 million TAM is my point. So put it differently, it's going to take many generations of products to really access the $1 billion TAM, and that's going to take a few years to get there. But I think we're very well positioned with our technology evolution.
Operator
Our final question is a follow-up from Tore Svanberg with Stifel.
Tore Egil Svanberg
Yes. I just had a 2-part question on the recent CommScope Vantiva deal. First of all, did that also have an impact on sort of purchasing behavior? Obviously, when you have an event like this, perhaps customers a bit more careful about buying inventory.
And then the second part of the question, does this change anything at all for MaxLinear? And the reason I'm asking that question is because now that it's sort of like 2 customers and 1 I would think that the qual for DOCSIS 4.0 is going to be a bit more simpler. So if you could answer those 2 questions, that would be great.
Kishore Seendripu
So Tore, your first part of the question is very, very easy. I don't think anybody knows which deal happens when, it's a very nonlinear process, M&A activity. Having said that, there's so much inventory in the channel. I don't think this -- the deal itself had any impact on that. It's really driven by the end market throughput and the end market itself has got a lot of inventory sitting on it, right? So I think our customers are in the same bad place we are in, number one.
Number two, regarding the DOCSIS 4.0 cycle, Yes, we've got a great chip, the lowest power, the most beautiful thing in the world. That's sort of a chip we have compared to any competition out there. However, we have seen the way the DOCSIS cable market plays out, it really takes 3 to 4 years to get to the ramp point where it hit 50% of the volume, at least 4 years to get to 50% of the market. And then a 7-year life to it, right?
Having said that, DOCSIS 4.0 is not the same for everybody, there is (inaudible) in the marketplace. There are 2 flavors of it, and it's a very, very costly network rollout. And I think operators are going to be very, very selective about DOCSIS 4.0. It's going to be a very small share of the market.
However, DOCSIS 3.1 has got many flavors. There's something called the Ultra DOCSIS 3.1, which is going to meet all the category of services that the market needs with its 10 gigabit received bandwidth and upstream multi gigabit, it's going to meet all those things. It's called the higher split Ultra DOCSIS 3.1. And that can roll out today based on all the network out there. And I think that's what the operators are going to lean towards, and that's going to be 80% of the market. And so DOCSIS 4.0 is great, but 3.1 is even more beautiful, Ultra DOCSIS 3.1 is what I put my bet on.
Operator
I was just going to hand it back over to Dr. Kishore Seendripu for closing comments.
Kishore Seendripu
Well, thank you -- thank you, operator. So I just want to thank you all for attending this call. I would like to tell you that we'll be participating in a number of conferences in November through January. The Stifel Midwest Growth Conference in Chicago on November 9, the ROTH Capital Technology event in New York on November 15, the UBS Technology Conference in Scottsdale on November 28, the Wells Fargo TMT Summit in Rancho Palos Verdes, California November 29, and the Needham Growth Conference in New York on January 18. With that, I want to want to thank you all once again for joining us, and we look forward to speaking with you again soon. Thank you.
Operator
Thank you. This will conclude today's conference. You may disconnect your lines at this time, and thank you for your participation.
