SMTC - NASDAQ

Greetings. Welcome to Semtech's Q1 FY '25 Earnings Call. At this time, all participants are in a listen-only mode. A question-and-answer session will follow the formal presentation. [Operator Instructions] Please note, this conference is being recorded. I will now turn the conference over to your host, Mark Lin. You may begin.

Thank you, operator. Good day, everyone, and welcome. This is Mark Lin, Executive Vice President and Chief Financial Officer. And I'm joined today by Paul Pickle, President and Chief Executive Officer. Today, after market close, we released our unaudited results for the first quarter of fiscal year 2025, which are posted along with an earnings call presentation to our investor website at investors.semtech.com. Today's call will include various remarks about future expectations, plans and prospects, which comprise forward-looking statements. Actual results may differ materially from the results anticipated in these statements due to various factors included in today's press release and presentation as well as in the Risk Factors section of our annual report on Form 10-K for the year ended January 28, 2024, filed with the Securities and Exchange Commission. As a reminder, comments made on today's call are current only as of today and Semtech undertakes no obligation to update the information from this call should facts or circumstances change. Unless otherwise noted, all income statement related financial measures will be non-GAAP other than net sales. Please refer to today's press release and presentation materials posted at the Investor Relations section of our website for information regarding our non-GAAP financial results and a reconciliation of our GAAP and non-GAAP financial measures. With that, I'll turn the call over to Paul to discuss our business and end markets.

Thank you, Mark. I'm pleased with Semtech's solid first quarter financial performance, with net sales above the high end of guidance, along with meaningful declines in channel inventories across each of our end markets. Now, moving to our first quarter end market results, starting with Infrastructure. For the first quarter, Infrastructure net sales were $56 million, a sequential increase of 42% and up 44% year-over-year. Results were at or above our expectations across each application. POS trajectory was similarly encouraging, up 33% sequentially and up 80% year-over-year. Channel inventories have largely normalized and are down 8% sequentially and down 18% year-over-year. Net sales for data center were $21.2 million, up 20% sequentially and up 61% year-over-year. In hyperscale data center applications, net sales more than doubled over last year with powerful secular trends in AI continuing to support our copper and optical portfolios. We believe AI is a trailblazer, which facilitates accelerated adoption of leading-edge technologies such as active copper cables, linear pluggable optics, and 1.6T optical modules in general compute applications. For our copper edge linear redrivers, our expectations for timing and market opportunity in 1.6T active copper cables remain consistent with last quarter. We are closely collaborating in the qualification process, which is largely progressing to plan. While qualifications are currently ongoing, cable suppliers have received purchase orders for designs specifying our chips, which we believe to be a good indication of progress. Accordingly, we reiterate our expectations of shipments beginning for these designs at the end of fiscal '25 with a substantive ramp starting in FY '26. We remain very excited by ACC opportunities and are also constructively engaged on a host of key projects as system vendors, ranging from retimed 50-gig solutions for AI and general compute applications, which we expect to substantially contribute to current year net sales, to retimed 100-gig solutions, which we expect to ramp this year, and retimed 200-gig solutions, which we expect to ramp starting in FY '26. We believe these opportunities equate to a very rewarding market over the next few years. FiberEdge TIAs and laser drivers for 400-gig and 800-gig optical modules continued to ramp in the first quarter. We also have FiberEdge design opportunities for 1.6T optical modules with sampling expected in early FY '26 and a ramp starting in the second half of FY '26. Design win activity and bookings increased across the number of customers, and we are also pleased with multiple 100-gig per lane design wins with Tier-1 and Tier-2 customers. We expect growth in this application to continue in the second quarter. Our Tri-Edge products are also performing well, with strong demand from key hyperscalers for 50-gig PAM4 and 200-gig and 400-gig AOCs for AI and cloud applications. Our 50-gig Tri-Edge products offering lower power and cost have consistently maintained strong share versus DSPs. In LPO, we remain engaged with several key partners in accelerating adoption of this technology. Net sales for passive optical network products were above our expectations at $27.2 million for the first quarter, a sequential increase of 88% and a year-over-year increase of 64%. The drivers of the strong results are consistent with what we communicated last quarter following the release of tenders. Robust demand for our XGPON and XGSPON reflect our first-to-market position, technical leadership and superior product performance. Expected market drivers in the North American and EMEA markets are consistent with prior quarter, and we are pleased with our continued engagement with several large network providers in both 10-gig and 50-gig applications. As there was a bit of a pent-up demand reflected in Q1, we expect net sales of PON to moderate in Q2. That said, based on our current analysis of booking POS trends, we expect solid results for PON throughout the year. Regarding other products in the Infrastructure end market, wireless demand has been consistent, but at low levels. That said, we are well positioned to lead 5G-Advanced front haul deployments with our Tri-Edge and FiberEdge 50-gig wireless platform and through our active participation in the Mobile Optical Pluggables Alliance, which includes our key partners Ericsson and Nokia. Our 50-gig front haul systems are in qualification, and we expect initial sales in the latter half of FY '25 for these products and production ramp in FY '26. For the first quarter, High-End Consumer net sales were $34.5 million, a sequential increase of 8% and up 60% year-over-year. POS was seasonally flat quarter-over-quarter and up 24% year-over-year. Consistent with our expectations, channel inventories continue to improve, down 11% sequentially and down 22% year-over-year. Net sales in consumer TVS grew within expectations to $24.9 million, up 20% sequentially and up 108% year-over-year and paired with channel inventory decline. We believe our market share at several leading consumer products companies continued to grow with design wins including handsets, wearables, and tablets. Our performance has been driven by customer preference of our first-to-market high-performance protection devices. Our demonstrated capabilities to reliably deliver to customer demand while maintaining extremely high levels of quality has enabled us to secure a top-tier scorecard rating at our largest consumer electronic customer. We expect these capabilities along with market-leading innovation allow us to meaningfully outperform market growth. We also continue to extend the frontiers of our capabilities through the deployment of innovative power management solutions that effectively combine over-voltage, over-current, over-temperature and surge protection with traditional IEC ESD protection. We saw increased levels of interest and adoption for these products at key customers during the first quarter. Our class-leading PerSe proximity sensing products had healthy consumption, up 19% sequentially and up 62% year-over-year. One of the consumption drivers was the rollout of specific absorption rate standards that took effect earlier this year. This demand has been supplemented by PerSe design wins in several markets, covering tablets, earbuds, notebook, computers and smartphones. We expect the consumption trend to strengthen, driven by widespread adoption of features such as gesture-control for earbuds and gesture-controlled smart glasses featuring bone conduction speakers. In line with our expectations, channel inventories for proximity sensing declined 26% sequentially and 5% year-over-year. For the first quarter, Industrial net sales were $115.6 million, down 5% sequentially and within expectations. Our IoT systems business recorded first quarter net sales of $48.4 million, down to 26% sequentially and down 57% year-over-year. Consistent with guidance, shipments for this business reflect our desire for healthier channel and end customer inventories. That said, this business is showing signs of recovery with first quarter bookings up 47% sequentially. For our router business, bookings more than doubled both sequentially and year-over-year. Router bookings were skewed towards our higher gross margin products, but we saw gains across all product families. Continuing its very strong launch pipeline and bookings from the XR60, the world's smallest rugged 5G router, continue to grow. We are pleased that we completed major network operator certifications on schedule and have commenced initial XR60 shipments for commercial programs, with second quarter shipments expected to sequentially double. XR60 pipeline expected to close in the current year is also very strong, with opportunities in government applications serving border security, public safety and logistics, each have benefited from passage of a US Federal budget. XR60 pipeline has also broadly increased in utility, transit, and medical applications. Lastly, routers saw channel inventories decline 27% sequentially and 47% year-over-year. Bookings in our module business were up 22% sequentially with growth driven by our 5G offerings and supported by further certifications at global network operators. Module bookings were particularly focused in enterprise networking applications, mirroring market growth. Pipeline also grew across a number of IoT markets, including payment processing, energy and fleet management. We are also very pleased to have announced along with key partners satellite non-terrestrial network support in our LPWA modules, further advancing our position in fleet and asset tracking. We believe the hardware business reached bedrock in the first quarter with stability in the second quarter and growth in the second half of FY '25. First quarter net sales for our connected services businesses were $24.1 million, effectively flat quarter-over-quarter and within expectations for this relatively stable stream of recurring revenue. Net sales of our RF industrial products, including LoRa-enabled solutions, increased 76% sequentially and 19% year-over-year. POS increased 25% sequentially and 16% year-over-year. The first quarter was characterized by further strengthening of LoRa adoption in both private and public LoRaWAN networks. In addition to traditional utility use cases, we saw greater momentum for our LoRa solutions targeting application and connected spaces, including building maintenance and energy management. We are also seeing increased opportunities in city management and citywide smart lighting control. In order to further simplify network deployment and reduce system cost, we are pleased to have introduced two technical enablers to the market, LoRa relay and the LoRa single channel hub. The hub was launched at the Embedded World Show held in April with demos of this product being co-hosted by several Semtech partner companies. We see this product as an important enabler for the adoption of LoRa technology into smart home applications. For TVS products in this end market, we continue to focus on our strategy to leverage customer engagement and adoption of best-in-class consumer products to capture meaningful SAM. In the first quarter, bookings for these products encouragingly increased 61% sequentially with sequential net sales growth of 7%. Design wins reaching production in the first quarter span applications, including intelligent vehicle cockpits and advanced displays, vehicle antenna, medical equipment, industrial power over Ethernet, and body-worn cameras used by first responders. Now, I'll turn the call back over to Mark.

Thank you, Paul. For the first quarter, we recorded net sales of $206.1 million, up 7% sequentially and above the high end of our guidance range, with the Infrastructure end market contributing to the favorable results. Coupling net sales and channel inventory by end market and on a sequential basis, Infrastructure net sales were up 42% and channel inventory was down 8%; High-End Consumer net sales were up 8% and channel inventory was down 11%; Industrial net sales were down 5% and channel inventory was down 17%. Gross margin was 49.8%, up 90 basis points sequentially and up 130 basis points year-over-year, reflecting favorable mix and cost controlled overhead spending. Operating expenses were $77.4 million and at the low end of guidance. We continue to closely evaluate spending, and on a year-over-year basis, operating expenses declined $15.3 million or 17%. We believe current spending levels reflect a prudent level of cost control coupled with improved allocation of spending to drive near-term financial results. We are wholly committed to programs supporting customer projects. Operating income was $25.2 million and operating margin was 12.2%, a sequential increase of 300 basis points and a year-over-year increase of 290 basis points. Net interest expense was $20.5 million, in line with guidance. We recorded net earnings per share of $0.06 based on a diluted share count of 67.6 million shares. Adjusted EBITDA for the first quarter was $33.1 million and adjusted EBITDA margin was 16.1%. This compares to fourth quarter figures of $24 million and 12.5%, and prior-year first quarter figures of $30.8 million and 13%. Moving to the balance sheet, we ended the first quarter with a cash balance of $126.8 million. Working capital changes largely corresponded to revenue and cost of goods sold. Inventories nominally increased $3.5 million or 2% sequentially, in part to support second quarter shipments and to carry a nominal amount of wafer bank supporting active copper cable opportunities. Inventories are down 30% year-over-year. Principal outstanding on our debt was $1.4 billion, unchanged from the fourth quarter, with a weighted average interest rate of 5.86%. At the end of our first quarter, our consolidated net leverage ratio calculated in accordance with our credit facility was 9.5, and we expect to maintain compliance with our debt covenants for the next 12 months. Free cash flow for the first quarter was $1.4 million use of cash, reflective of working capital changes and we did not draw on our revolver. Now, turning to second quarter guidance, we currently expect net sales of $212 million, plus or minus $5 million. We continue to evaluate shipments into channel to maximize gross margin opportunities. We expect net sales from the Infrastructure end market to increase sequentially with data center applications leading the growth. As Paul mentioned, in the first quarter, net sales of PON were up 88% sequentially and we expect a bit of normalization for this application. We expect net sales from the High-End Consumer market to be slightly up with expected seasonality benefiting this market at the end of the second quarter through Q3. We expect Industrial net sales to be flat to slightly up, with bookings activity indicating a recovery in the second half of FY '25. Based on expected product mix and net sales levels, gross margin is expected to be 50%, plus or minus 50 basis points. Operating expenses are expected to be $77.5 million, plus or minus $1 million. We expect net interest expense to be $20.5 million, and a non-GAAP tax rate of 15%. These amounts are expected to result in a net income per share at $0.09, plus or minus $0.03, and adjusted EBITDA of $36.3 million, plus or minus $2.6 million. I'd now like to turn the call back over to the operator for Q&A.

Thank you. [Operator Instructions] Our first question comes from the line of Harsh Kumar with Piper Sandler. Please proceed with your question.

Yeah. Hey, guys. First of all, congratulations on what appears to be a pretty good turnaround that you've instituted already. Paul, probably, the question that I get most often from investors is on the active copper cable opportunity. There seems to be a little bit of confusion whether you're playing in the vertical aspect or the horizontal aspect or both. Maybe you could clarify that and maybe help us understand if it's the same chipset that you will use for vertical and horizontal applications? Just a little bit more color on this opportunity?

Yeah. So, we're playing in both -- and so every rack will have two sets of active copper cables within them. That would be a total of 36 cables, but it's not the 5,000 direct attached cables you'd have to subtract from that. But horizontal cabling in between the racks is a larger opportunity, but of course, really does depend on the build configuration or the shipping configuration up to 162 cables rather for those horizontal NVLinks. In terms of chipset, it's almost the same chipsets, but there's a few ICs.

I got you. Mark -- I'm sorry, Paul, for my follow-up, when I look at turnarounds, it's almost always the same sort of playbook; management comes in, fixes the business, cuts the cost, stabilizes the cash flow, and then the next steps are taken. You've done all these steps. You're kind of at a point where your business is self-sustaining, but there's also that question in your case of potential de-levers. And I was curious if you're at a point where you're starting to think about that opportunity set that's available potentially to you? Or is it too early to talk about it?

Well, I'll kick this question over to Mark, if you don't mind, Harsh.

Yeah. Harsh, thank you for recognizing in our Q1 results that OpEx has stabilized. And based on our outlook, we really do expect an upswing in cash flows and EBITDA growth to improve our leverage metrics. That said, consistent with what Paul said in his first earnings call with Semtech, a reduction in the quantum of debt beyond free cash flow generation is something we all believe is important here. I'll also add that we evaluate capital structure alternatives in the ordinary course, including a refinancing of our debt in addition to potential asset sales at, what I would consider, appropriate values. I covered our covenant compliance outlook in my prepared remarks. And others know that any disposition will not be at a fire sale and we'll need to appropriately value the asset. You can understand we can't provide timelines or specifics, but transaction proceeds we'd use to reduce debt and improving leverage ratio and financial metrics are all factors that we consider in a potential divestiture.

Yeah. And if I could summarize that, I'd say all tools are at our disposal, all options on the table. We definitely are focused on reducing the overall debt load and we'll take a look at every option available to us.

I appreciate it, guys. I wasn't looking for timing. I know you can't give that, but I appreciate the fact that that option is still open at some point in time. Thanks, fellas. Thank you.

Thanks, Harsh.

Thank you. Our next question comes from the line of Tore Svanberg with Stifel. Please proceed with your question.

Yes. Thank you, and congratulations on the continuous progress here. Back to the question on ACC, I think for 1.6T, Paul, you mentioned that the cable manufacturers are seeing POs and that you are qualified there. Could you just elaborate a little bit on that? And how should we think about timing there as far as when you would see that potential ramp?

Yeah, all right. So, as is kind of typical in hyperscale data center, we'll end up working with a lot of those hyperscalers as well as technology partners to spec product and build chips, collaborating on those chip definitions. As is often the case, they'll specify to us who they want us to utilize for fulfillment. A lot of times that does go to an active optical cable manufacturer, in this case, an active copper cable manufacturer. I'm not at liberty to say who those partners are, but there's two partners that have been designated. We're working with those partners for final qualification. And then, in terms of timing, it's still right now the timing is holding consistent with what we've said in the past, so the latter half of this fiscal year. And I would really kind of say that the ramp is next fiscal year, but we'd expect to see some early revenue this year.

Very good. Thanks. And as my follow-up for, Mark, so the way I understand this now, Mark, is that the OpEx is now sort of at the right level. And going forward, it would grow at half the rate of sales. Is that how should we think about it? Or could there still be some further OpEx reductions this fiscal year?

Yeah. I'm pretty comfortable with where the OpEx levels are right now, but we are constantly looking at how to better spend the dollars, better allocate the dollars for near term. As revenues grow, there's going to be some OpEx growth, just within sales and marketing that can be expected. But the growth rate that you provided, I can sign off on that.

That's very helpful. Thank you. Congrats again.

Thank you.

Thank you. Our next question comes from the line of Quinn Bolton with Needham & Company. Please proceed with your question.

Hey, guys. I'll offer my congratulations as well. Just wanted to follow-up on the ACC opportunity. I know you're going to get a lot of questions here, but there's been a pretty wide range of expectations for this opportunity for you guys. And wondering if you might want to try to level set some of those expectations. I mean, how big -- what do you think a reasonable expectation might be for this business, say, in fiscal '26? I know you'll get a little bit of revenue maybe preproduction in fiscal '25, but as we look out to next year, how big could this be?

So, rather than answer how big could it be, because to imagine the high end, it's a pretty big number, I would prefer to kind of focus on what I think is currently a bit more of a reasonable expectation, at least what we're kind of planning around. So, the number of cables that could be used is heavily dependent on both the rack configuration and NVL72 versus NVL36, and the number of horizontal connections, and obviously, the number of NVLs that we'll ship next year. So, it's heavy dependence on shipment configuration, but to just cut to the chase, we kind of size it at $100 million opportunity, not as base case, I definitely don't want to put a high side case number out there. I think on that base case, it's reasonable to expect that we're going to share production between us and one other component supplier. And if you want to just put a slug in there for the share that we would see, you could call it 50-50. We were first to provide the chips, first to get designed in. Maybe that gives us a slight first-mover advantage, but I think it's reasonable to expect this kind of settles out down to a 50% share split.

Got it. A quick clarification there, Paul. So, that $100 million, is that the opportunity for the total TAM that would then be split by you and the competitor, or is $100 million sort of the expectation for Semtech revenue? And then, I've got hopefully a follow-up.

So, that is total SAM and then we would get a share of that. And I'm looking at a particular program use case. So, two models, different configurations, but just one particular program.

Got it. Okay. Thank you for that. And then, you'd mentioned LPOs, some of the DSP vendors are suggesting that LPO activity may have slowed as you get the linear receive DSPs coming to market. Just wondering if you could sort of share your thoughts. Where are you seeing interest in LPOs? Do you see it in GPU back-end networks? Do you see it in general switch infrastructure and data centers? Are there other use cases where you're seeing interest in LPOs growing?

Yeah. So, all right, I guess if I was a DSP vendor, I'd probably downplay it as well. But there is -- the jury is still out in terms of how widespread. GPU-based back-end networks -- you got a GPU vendor out there that has already made the statement that they'll end up using LPO. That's a significant SAM in and of itself. And whether or not it ends up in general data center use in the fiber map, I think that's where some of the question might lie. And I will say the jury is still out. It does kind of come down to operation over all parameters. And right now, it's still kind of hinging on lifetime data. We think it looks good. The one thing that I know for certain, the motivation there to use it and implement it is quite high. It's one-fifth -- let's assume that nobody cares about the cost of the DSP, which I think is probably the case. It's not about saving money, but it's one-fifth the power consumption over full retime solution. So, you've got a couple of variances. You could end up with full retime solution, a half retimed or LCO solution, and then a linear pluggable solution. I think either way, it shakes out. Right now, data center is going to go through a boom for the next three to five years. There's going to be lots of variations of technology implementations. And it's all going to come down to who can squeeze out the best performance for the power limitations that they have. And so that's why we're still rather bullish on the opportunity.

Perfect. Thank you, Paul. Thank you, Mark.

I'll tag on one little quick answer to that answer I just gave. If you look at the really kind of compelling reason to use ACC and the reason why copper is coming back in terms of popularity is because if we look at a DSP-based solution versus copper solution as in an AEC versus an ACC, it's one-fifth the power. So, it's really compelling, works up pretty well in short reach, up to so many meters. But when you go beyond that, you're going to need retime solutions. So, I think everybody is going to have a play over the next three to five years.

Thank you.

Thank you. Our next question comes from the line of Rick Schafer with Oppenheimer & Company. Please proceed with your question.

Thanks. And I'll add my congratulations. And, Paul, just on that last comment really quick, kind of a follow-up question. I mean, does that mean -- do you guys have the capability or the desire? Would you guys move into AEC to kind of get maybe move a little bit longer reach than say the 2-meters or wherever ACC is right now?

I'll put it this way. I'm a big fan of land and expand. So, if you've got a good profitable relationship with the customer, you really need to start looking around at other sockets that might be in the realm of possibility. I don't like it when companies take technology-specific stand or they kind of -- I prefer a bit of more of a technology-agnostic approach. And so, this is an area where over the next five years, I think we're going to be focusing pretty hard on the strategy and looking to fill out some of those needs that we think are going to come in four to five years, different architectures for retimed data transport solutions is going to be a necessity of ours and we're going to continue to look to expand the product portfolio.

Thanks for that. And if I could just ask a quick one on Tri-Edge. You mentioned it sounds like it was -- it had picked up in the quarter, if I read that correctly in your comments? And I just was -- I was sort of trying to see if we could get a better frame out of what contribution you expect this year or what that SAM looks like today for you guys versus, say, a couple of years out. I guess I'm really just trying to understand the opportunity set here and where -- I understand it's lower power and lower cost and all that, I mean, where is it going to play, your analog panel, excuse me?

Anytime somebody can take an analog approach, it's great. The disadvantages to it is it does require tuning. We've got some software that we wrap around it to tune the solutions. And so, it really kind of -- in a world where everybody is counting picajoules per bit, I think if you can make it work with analog, it's great. Analog also has some advantages in terms of latency. Let's say general compute data center doesn't really care about latency, but AI certainly does. If you're talking about linking GPUs, NVLink really can't tolerate latency. And then, we've also got some really strong positions at Nokia and Ericsson in telecom front haul because they can't introduce latency either. So, there's definitely a play there. I think it's not the end all, nor is the DSP-based solutions. Each one's got its advantages. But for us right now, Tri-Edge is just kind of if we look at it on a go-forward basis, just over the next quarter, we would expect it to significantly increase and perhaps double. So, we're just seeing a nice strong uptick for data center overall.

Okay. Thanks a lot. I appreciate it, Paul.

Thank you. Our next question comes from the line of Cody Acree with The Benchmark Company. Please proceed with your question.

Yeah. Thanks for the help. Paul, can you maybe -- you gave us the details on active copper cables. Can you maybe provide some of the similar framework around your expectations for LPOs?

Yes. So, right now, we're kind of -- we've got a design project with a particular, what I would call, technology partner. It's the same partner around ACC, specifically for LPO. Most other vendors, we've got a tech partner in Broadcom also that we're testing out solutions. We're going through lifetime testing with another OEM. So, it's making its progress. And then, I think as we kind of look at some of those LPO products, they're actually being teamed up with in-house DSP solutions from other technology partners that are not meant as a standard product portfolio per se. So, good activity there across the board. So for us, we don't need it to be LPO necessarily. These 200-gig per lane single lambda parts are going to find their way to a lot of different applications, whether they're just really good strong linear TIAs and drivers, and that kind of lends itself to being in a direct drive or in a half retimed or in a full retimed, really doesn't matter to us. I think we're going to do rather well with the product line.

So, what are your expectations around LPOs for next year?

Not currently. I think we've got -- we're a little bit further along the design curve, the design in and win curve with ACCs. I think I'll be able to put better color to our 200-gig single lambda product and 1.6T optical solutions in the next couple of quarters.

Okay. And then just lastly, can you just help us with the migration of the industry to 200-gig single lambda? Just where you're seeing the breadth of implementation today and where do you expect that to go?

Yeah, that's kind of interesting. We're talking about 100-gig, we're talking about 200-gig, but the reality is, we still -- the majority of our revenue in FY '24 was on 50-gig per channel. And even today we're starting to leg into 100-gig and it's getting an appreciable amount. Let's say it's doing at least as much as 50-gig. But I think 50-gig per lane is still going to be the bulk of it. So, this is a bit more of a -- in terms of the total shipping opportunity, I think this year is really going to be the transition year where we see general compute kind of migrate towards 100-gig solutions. That's still the expectation, although it's moving fairly slowly, not at a rapid pace. But AI is really kind of driving almost skipping a major note here at 100-gig moving straight to 200. So, it's a little hard to say what that transition in terms of revenue contribution is going to be like right now.

All right, great. Thanks, guys.

Thank you. Our next question comes from the line of Tristan Gerra with Baird. Please proceed with your question.

Hi, good afternoon. So, you've talked about ACC potentially moving to general purpose data center in '26. In AI, how sustainable is that in future AI platform beyond this year? When do we see horizontal and vertical connection moving ultimately to optical? And if not, why not?

It could very well be, but -- all right, so I'll tackle that last bit, Tristan, if you don't mind. If we look at configuration [indiscernible] are going to be close together and the reach is rather short. It doesn't -- you don't necessarily need optical for that. Where would it make sense to do optical? Well, if we can't do-400 gig, single-lane 400-gig active copper cables, well then I think the natural migration path would be to optical for up to some reach. But that's where you have to kind of think in terms of internal to the rack, those direct attached cables, are they going to have to go active versus passive solutions? At some speed, a lot of people would say, yes. I think we have to kind of wait and see how that goes. It really kind of depends on the switch ASICs and how strong their drivers are. But I don't see active copper. If the question is, is active copper going to be something that's short term? I think the answer to that is absolutely not. It's going to be another tool in the arsenal and it all kind of comes down to reach. But if you look at those horizontal cables, if we can't do 400-gig active copper at, let's say 1.5 meters or so, then direct drive, no latency, optical connections is the next logical choice rather than retime because you can't introduce latency into those NVLinks. Does that answer your question?

Yes, very helpful. Thank you. My follow-up question is, looking at the gross margin catalyst for the second half of this year, is that just a function that's optical growing faster than the average, or should we be looking at other catalysts in terms of continued gross margin upside?

Well, there's a few catalysts. So, we've got -- if you think in terms of some of the end markets that we said are fairly soft. Telecom certainly is a positive gross margin driver when that comes back. PON has been a decent driver as well. LoRa is a pretty decent margin for us, so that would be a good driver for us. Anything except for -- the hardware business, the module business tends to be a little bit lower, quite a bit lower. And then that routers business mid-50s or so is typically what it is. But any of those trends, I think are going to really give us a boost. Mark, did you want to add anything to that?

Yeah. It's largely a matter of mix, but we see some of the faster growing businesses at the higher margin range, offset by -- a bit by the hardware business recovery in the second half.

Yeah. And consumer goods, it's been fairly strong for us. It's still we see support for consumer goods that has a tendency to be a little bit lower, a little bit more dilutive to corporate margins. So, if that backs off and something else fills its spot, you're going to see that gross margin go up.

Great. Thank you very much.

Thank you. Our next question comes from the line of Craig Ellis with B. Riley Securities. Please proceed with your question.

Yeah. Thanks for taking the questions, guys, and congratulations on the progress. So, I was hoping, Paul, if we could just go back to some of the comments on active copper cable, and at the risk of getting the cart too far before the horse, if we think the opportunity may be $100 million opportunity in fiscal '26 that you'd be splitting, how do we think about the growth rate beyond that? I know a number of companies have put out chip-related AI growth rates that would be in the 30% to 50% CAGR range. Is that the range that this business could grow in or how do we think about what happens after fiscal '26?

I do think that it could grow a multiple of those chip rates. So, that 30% to 50%, I think can be higher, but it really does -- the secular trend that we're going to look to is that shipping configuration. If people are actually putting these in multi-rack installations to take advantage of multiple racks being grouped together in giant GPU clusters, and that's largely dependent on use case. Use case is kind of taking off favoring a massive parallel architecture, then the opportunity I think is -- it could be quite explosive. So, it really kind of depends on what that shipping configuration looks like. And as AI continues to evolve, how they want to utilize available exaflops.

Got it. And that's really helpful. And then just going back to some of the other comments and thinking on a multiyear basis about Signal Integrity in total. So, we've got the base station business that's been pretty quiet, but bouncing along the bottom. China PON taking a breather this quarter, but has room to at least reaccelerate a bit. And then data center and its AI growth component looked pretty strong. So, as we look out beyond fiscal '25 to fiscal '26, it seems that all of your segments should be contributing to growth, or is there something that you see beyond this year within Signal Integrity that would cause one of the four big drivers to take a breather?

I don't see the drivers taking a breather. One of the first things I did when I got here, I kind of looked at 10-year trends for some of the product lines. And Signal Integrity is one of those that grew very consistently, albeit had some definite cyclicality to it. But if you draw a line through it, it's an 8% CAGR, and that's before AI became an upside SAM opportunity for us. So, I think it's going to uptick from there. I think it's going to have some pretty decent multiple year growth -- sequential growth. So, we should be in pretty good shape at this moment.

Got it. Thanks for the help. Good luck, guys.

Thank you.

Thank you. Our next question comes from the line of Scott Searle with ROTH Capital Partners. Please proceed with your question.

Hey, good afternoon. Thanks for taking the questions. And nice job on the quarter. Nice to see the recovery moving back in the right direction on the semi front.

Thank you, Scott.

Hey, Paul, maybe moving away from the optical side of the equation for a second, to follow-up on the prior question on PON, it sounds like this quarter there's a little bit of digestion. I'm wondering if you could address the drivers. Are they specifically China right now? And have we seen a watermark? And as it relates to LoRa, it also had a similarly strong quarter. What's the long-term growth rate that you're thinking about for the LoRa opportunity going forward?

Okay. Scott, we had a little bit of a hiccup in your audio. What was the first question again?

Oh, my apologies. Just, PON, in terms of drivers, is it moving beyond China? And is it $27 million that you saw this quarter, is that kind of a high watermark for the next several quarters, or is that something we're going to breach as we get into fiscal '26? And on the lower front, when you came in, I think you recalibrated expectations of the long-term growth rate in that marketplace, but a nice recovery this quarter. I'm assuming that some of that is just a little bit of inventory backfill and channel demand snapping back a little bit. What is the long-term growth rate that you're thinking about for LoRa going forward?

Okay. So, I'll take your first question first. So in terms of PON, definitely China is behind this kind of recent uptick. If we look at PON, it was definitely down. We had a delay in PON China tenders. But overall, I think there's really some underlying strong trends there in terms of PON. I wouldn't -- when we start talking about China PON business, everybody kind of equates it to that real estate market, knowing it's down, but this is more of an infrastructure upgrade. So right now, we've had a shift. We're about 50% of the number is coming from 10-gig. So, this is a really strong indicator that we're in a leadership position there. It is an upgrade cycle. It's going to be a very profitable cycle for us. And I would expect a little bit of peaks and valleys between the quarter. But overall, we're going to see -- I'm expecting growth in PON and that trend kind of continuing. So right now, I don't have a reason to think that it's going to stop. In terms of high watermark, I think we're probably at a good number for the moment, the $27 million. We could -- it could edge up higher, but I'm not really willing to kind of count on that at the moment. I think we're in healthy territory. We'll see -- there's some order hesitancy and lack of visibility as the tenders rolled out. So, we kind of have to see that kind of stabilize. And we did have a strong catch-up quarter, where we had a big boom of bookings in January, and we really had to scramble hard to kind of ship that product. And so that's why it pulls back a little bit as we catch up. But we're on a ramp cycle, expect it to continue, so good growth. In terms of other markets, I would point to the broadband equity access and deployment initiatives out of North America or BEAD, as they call it. I don't know who picks these acronyms, but it's basically a government subsidy to roll out communications infrastructure in urban markets. So, this is going to be pretty good opportunity for us. Fiber is kind of the way to go. Copper is running out of gas and people don't really want to share those copper pipes. When everybody gets home, you can have 1-gig worth of access and you're still turning 200-meg and so that's kind of -- that's got to change. I apologize, I said urban markets, it's rural markets. So, this is an infrastructure build out that's coming from US government funds. So, I think we've got some good trends happening with PON. It's not going to change anytime soon. And in terms of lower growth, I would say double digits, easy. That's my expectations. However, we want to see that business recover, and we really want to kind of fill out the strategy. We've got some nice development going, so we're going to support a bit more protocols with our next-generation chipsets. We'll be able to support Wi-SUN. You're going to see those announcements over the next couple of quarters as well. We're getting some really nice feedback on that. And so, this is the really kind of open up markets for us. People that don't really want to close protocol market, they can adopt an open one and still utilize the same chipsets from Semtech. And then, we've got some ideas on how to roll out some support tools, configuration tools, device management tools that are -- is going to make deployment a lot easier. So, expect to see some news along those lines over the next few quarters.

Great. Very helpful. And by the way, BEAD is one of the better acronyms out there, sadly. But a quick question on [indiscernible] modules. That -- look, it's been a lot of pressure. There's been a lot of channel inventory. But it sounds like the order trends are coming back. I think it's probably down about 70% from peak revenue levels. So, I'm wondering if you could calibrate us quickly in terms of what the point of sale looks like, how you kind of expect the recovery in the second half of this year? And what's a normalized number to think about as we start to see some of those trends picking back up? I wouldn't expect us to go back to where we were two years ago, but is there a normalized number that you're thinking about? Thanks.

Yes. So LPWA is actually leading it. So, we're actually seeing municipality rollouts and adoption upticking. And this is kind of interesting because if you look at the previous peak revenues, they were a little bit more heavily broadband, but we've got a lot of design ins on LPWA. And so, we're starting to see that adopted in metering applications. And so, expect that that end market to kind of recover a bit faster. It's still reported in industrial, but broadband revenue is down mostly because we had heavy, heavy buying from our end customers and not a lot of visibility, it's not channel inventory, it's end customer inventory. And so, as that bleeds down, they also had POS that dipped. So, our industrial POS dipped, their industrial POS dipped. And as a result, that channel inventory in terms of days of inventory outstanding kind of ballooned with both the numerator and the denominator changing. So, we're starting to see signs of life there. I still don't have a ton of visibility on it, but I think my remarks are still very consistent from last quarter. I was talking about Q1 being the bottom. I was hoping Q1 was the bottom, Q2 might be slightly lower or sideways. I think that's where at, Q1 being the bottom. We hit bedrock, as I said in my prepared remarks. Q2, slight uptick. And but we do see industrial POS and that end market coming back a little bit. So, we could get a little bit more of a snappier recovery in that hardware business and that's kind of what I expect. I did see signs of life in routers. And so our end customers of broadband modules should have already or should have also seen some signs of life there. So, I think right now we're kind moving exactly as we stated. So, second half, snap year recovery, probably a little bit slower in Q3 timeframe, but we'll have better visibility on those inventories. So, what's a normalized number for modules in particular? It's been as high as -- I want to say, 355. I'm doing that from memory. Maybe Mark can correct me. And if I go back to my November call, I think I gave everybody kind of a normalized revenue number of the ICS, routers and modules coming back to on the order of about a $460 million business. I've also said $400 million -- let's say, $460 million fully recovered, $400 million run rate exit this year, kind of driving towards. I still think we're probably there and modules can certainly get mid-2s somewhere in there. But I'll be able to give you a little bit more color as time goes on.

Very helpful. Thanks so much and congrats on the quarter again.

Thank you.

Thank you. And our next question comes from the line of Anthony Stoss with Craig-Hallum. Please proceed with your question.

Hi, guys. Nice execution. A lot of my questions have been asked, but Paul maybe I can follow-up on, do you expect any of your business segments to be down sequentially in July? And I guess maybe just a leading question on the October quarter, do you think revenues in every business segment should be growing especially now that you've mentioned the CRP should be growing?

So, I think in terms of segments down, we've kind of talked a little bit about some of the pausing that has a tendency to happen. We'll get a rush of product out there, it's got to get built. So, we're going to see some peaks and valleys on a quarter-by-quarter basis. Quite honestly, I wouldn't say a segment is down. We'll have results that will vary quarter-over-quarter, but I wouldn't say a segment is down at this point, because we're really kind of skipping off the bottom with a few segments coming back. So, telecom, the one thing I will say, telecom remains muted. We've got lots of good design activity out there, but I'm not going to -- I wouldn't actually start to get a bit more positive about it until I started to see some CapEx trends or budgeting trends in that space. And really that starts to happen when 5G advance starts to people start thinking about rolling that out. So right now that remains muted and I don't have any visibility at this point to report that would say that we're going to see a change anytime soon. But other than that, things are clipping along. LoRa as a revenue number is not quite there. We reported positive POS, so that's a really good trend. We are seeing we are expecting those numbers to come up, because it has been really off of a low, but I don't really have anything else that I would say is down.

Got it. And then just a quick follow-up on the ACC side. You talked about qualifications are ongoing. Any sense of how much longer it's going to take?

So, I think on my last call, I said we'd know in the next two quarters. We're reporting incremental progress with purchase orders at the cable manufacturers. I think we should hopefully find out in a quarter and have something to report.

Very good. Thanks, guys.

Thank you.

Thank you. And we have reached the end of the question-and-answer session. I'll now turn the call back over to Paul Pickle for closing remarks.

Thank you, Shalami. Appreciate you guys joining us today, and have a great evening.