Q3 2024 Semtech Corp Earnings Call
Participants
Mark Lin; Chief Financial Officer, Executive Vice President; Semtech Corp
Paul Pickle; President, Chief Executive Officer; Semtech Corp
Quinn Bolton; Analyst; Needham & Company
Tore Svanberg; Analyst; Stifel Nicolaus
Craig Ellis; Analyst; B. Riley Securities
Anthony Stoss; Analyst; Craig-Hallum Capital Group, LLC
Christopher Rolland; Analyst; Susquehanna Financial Group
Rick Schafer; Analyst; Oppenheimer
Tristan Gerra; Senior Research Analyst; Robert W Baird & Co Inc
Harsh Kumar; Analyst; Piper Sandler
Cody Acree; Analyst; The Benchmark Company, LLC
Scott Searle; Analyst; ROTH Capital Partners
Presentation
Operator
Good day, and thank you for standing by. Welcome to Semtech Corporation's third-quarter 2024 earnings conference call. (Operator Instructions)
Please be advised that today's conference is being recorded.
And I would now like to hand the conference over to Mark Lin, Executive Vice President and Chief Financial Officer. Thank you, sir. Please go ahead.
Mark Lin
Thank you, operator. Good day, everyone, and welcome to all those joining today's call, including analysts, investors, and my fellow employees. I am 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 third quarter of fiscal year 2024, which are posted to our Investor website at investors.semtech.com. Supplemental earnings materials, including net sales data by end market, reportable segment and geography, as well as a share count table reflecting potential share issuances from our convertible notes at various stock prices are also posted to our Investor website.
Unless otherwise noted, all income statement related financial measures will be non-GAAP other than net sales. A discussion of why the company considers such non-GAAP financial measures useful along with reconciliations of such non-GAAP measures to the most comparable GAAP financial measures are included in today's press release.
Today's call will include forward-looking statements that include risks and uncertainties that could cause actual results to differ materially from the results anticipated in these statements. For more detailed discussion of these risks and uncertainties, please review the Safe Harbor statement included in today's press release and in the Risk Factors section of our most recent periodic report 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.
With that, I will turn the call over to Paul to discuss our business and end markets.
Paul Pickle
Thank you, Mark. Semtech's Q3 financial performance generally met my expectations with net sales slightly above the midpoint of guidance. I'll provide a bit more color on end market performance, which we believe improves understanding and comparability of our business before turning the call back to Mark.
Infrastructure net sales were $43.2 million, (sic - see slide 3, "$43.1 million") a sequential increase of 2%. Hyperscale data center applications benefited from continued momentum of AI-driven applications and grew both sequentially and year over year.
We recorded growth in general compute data center applications as well. We had record shipments in 200 gig, 400 gig, and 800 gig applications with our Tri-Edge and FiberEdge products having been well received in data centers around the world. Net sales for other products in the infrastructure end market, including passive optical network, wireless, and infrastructure-focused transient voltage suppression faced a headwind from elevated channel inventories, though POS for each of these products grew sequentially.
As stated previously, PON tenders are expected in our fiscal fourth quarter, and we have seen some early channel pools in anticipation of this timing. We believe we continue to lead the PON space. Last March, we announced the world's first 50 gig PON chipset, which enables multi-gigabit broadband services for multi-tenant locations and will drive new use cases for the PON market, including enterprise applications for small business needs.
This chipset is currently sampling with module vendors and system integrators with positive feedback. We have also started sampling 1.6 terabit in linear pluggable optics applications at optical module vendors and system integrators with early success. We believe our technology solutions have meaningful advantage in AI applications, offering lower cost, lower power, and lower latency. Net sales from these leading-edge applications are expected to begin in the second half of calendar year 2024.
On the automotive front, while net sales are currently small, we believe we are making inroads with the next-generation Ethernet protection devices. Adoption of automotive Ethernet continues to accelerate, and our leading products in this space are positioning us for future growth.
For the fourth quarter, we expect net sales from the infrastructure end market to be flat to slightly down as the market further digests channel inventory. High end consumer net sales were $37.6 million, a sequential increase of 10% benefiting from seasonality and TVS products primarily directed at smartphone applications.
Sales from these TVS products increased both sequentially and year over year with corresponding increases in POS. Strength in the quarter was driven by measurably greater protection content and current-generation products at a leading North American smartphone manufacturer compared to prior generations. Design wins also grew 10% sequentially for handheld devices and wearables, further validating our recent innovations in this space. Sales from proximity-sensing products declined slightly on a sequential basis, but grew year over year. While encouragingly, POS grew on both a sequential and year-over-year basis.
Our PerSe sensing solution provides capacitive touch and user sensing capabilities. For wearables, PerSe sensors improve the user experience by providing automated functionality and rapid response interactions such as smart assistant activation, noise cancellation control, and media player management.
Our PerSe sensors also provide the industry's best performance of minimizing end users' exposure to RF energy. This allows equipment manufacturers to achieve compliance with specific absorption rate or SAR regulations while maintaining optimal signal connectivity.
While PerSe is well known in the smartphone space, efforts to reduce SAR have opened opportunities and design-ins for applications, including notebook computers, tablets, and wearables. For the fourth quarter, we expect net sales from the high-end consumer market to be down due to the aforementioned seasonality and channel inventory digestion.
Industrial net sales were $120.2 million, down 26% sequentially, but within expectations, with the bulk of the decrease in the ISP modules business. Semtech delivers product leadership in low-power wide area and broadband applications and we'll continue to leverage these areas to return the business to growth.
IoT connectivity services reported net sales of $24.2 million, up 1.4% sequentially, highlighting the stickiness and relative stability of this recurring revenue stream. Within AirVantage smart connectivity, we continue to see growth in our advanced service line, which offers global access through a multi-EMC, multi-profile, single sim solution.
Our carrier plus offering, which simplifies regional deployments with multiple Tier 1 carriers, is showing growth, particularly in Australia and New Zealand. These growth areas are partially offset by the sunsetting of legacy technologies, but net to stable sales.
Net sales of LoRa-enabled industrial products declined sequentially, but POS and bookings increased, providing some indication of market recovery. Connected gateways increased 3% sequentially, and we are encouraged by substantive increases in POS for in-node products. Adoption of LoRa-based solutions in the utility space gained further traction in Q3 with multiple European-based utilities announcing RFPs with a requirement for a LoRa solution. We are also integrating LoRa into custom SOCs for a customer, which has expanded opportunities in hearing aid applications.
Returning to our hardware business and expanding on the regulatory environment I discussed in our last quarter's earnings call, as the largest North American module supplier, Semtech is definitely seeing a meaningful share shift in pipeline from our competitors to us and pipeline is translating to design wins. Customers in this market have varying degrees of risk tolerance, most sensitive as critical infrastructure services such as industrialized laptops used by first responders and cellular routers used by utilities.
We are now benefiting from the next wave of concern with applications that involve payments or personal identifiable information. Point-of-sale devices are prime example, where there was perhaps a hyper focus on cost for point-of-sale applications, the security of Semtech as a North American supplier has resulted in design wins that translates to revenue in the second half of next year.
Our near-term headwind for our router business is government-spending constraints under the current federal budget situation with traditional end-of-year public sector spend lower than trend. For the fourth quarter, we expect industrial net sales to be flat to slightly down with continued inventory digestion and the aforementioned public sector headwinds, but stable net sales from our managed connectivity offerings.
Now I'll turn the call back over to Mark.
Mark Lin
Thank you, Paul. Turning to our Q3 results, recorded net sales of $200.9 million, slightly above the midpoint of the guidance. Paul discussed end market net sales performance with infrastructure up 2% sequentially, high-end consumer up 10% sequentially, and industrial down 26% sequentially. I'll refer participants to our investor website for the last five quarters of net sales data by end market reportable segment and geography.
Gross margin was 51.3%, a sequential increase of 170 basis points and above the high end of our guidance for the quarter. Gross margin reflected well negotiated supplier concessions to reduce third-quarter manufacturing costs, offset by incremental inventory reserves and unfavorable mix. Supplier agreements particularly benefited IoT systems for the third quarter. Netting these items, consolidated gross margin would have approximated the midpoint of guidance.
Operating expenses were $82.5 million, favorable to the midpoint of guidance, with R&D expense sequentially increasing about $900,000 and SG&A sequentially decreasing by $4.1 million. Net interest expense was $22.3 million, with only about one weeks benefit from the capital structure change completed at the end of October. We recorded net earnings per share of $0.02 based on a diluted share count of 64.3 million shares.
Changing gears to the balance sheet. We ended Q3 with a cash balance of $123.8 million and undrawn revolver capacity of $280 million. Working capital moved in a favorable direction with inventories decreasing $19.6 million. Principal outstanding on our debt was $1.4 billion with a weighted average interest rate of 5.57%.
At the end of our third quarter, our consolidated net leverage ratio calculated in accordance with our credit facility was 6.49, and we expect to maintain compliance with our debt covenants for the next 12 months. Including the effects of our convertible notes and interest rate swaps, loan principal was about 83% fixed and 17% floating at the end of Q3 and all of that loan principal was long term. At the end of Q3, there were no scheduled principal payments on our senior secured credit facility until January 2028 and no scheduled principal payments on the convertible debt until November 2027.
I'd like to discuss the five-year $200 million, 4% convertible note we issued this past October. In addition to the benefits of increasing our mix of fixed to floating rate debt and eliminating scheduled principal payments on our term loan, we placed emphasis on reducing cash interest costs. Our Federal Reserve rollback in interest rates may occur. When I turn on the high beams, I expect to be in a protracted elevated interest rate environment.
As I mentioned, all of our loan principal is long term. While deleveraging is a key use of excess cash to convert eliminated $67 million in scheduled principal payments next fiscal year, we're adding optionality. Also addressing dilution, please refer to the convertible notes dilution table posted at the Semtech investor website. This table provides incremental dilutive shares on both the 2027 and 2028 convertible notes.
On both the notes, par value or principal is paid in cash and conversion of par is settled in cash or shares at Semtech's discretion. Assuming the conversion of both notes were settled in shares, dilution of 20% is reached at a stock price a bit above $85 with no effect on diluted shares today.
On the date we priced the 2028 notes offering, a $250 million raise would be the equivalent of 15.6 million shares or an immediate 24% dilution, which I do not believe counteracts the benefit of reduced interest expense. I hope this extended explanation provides some insight into management's calculus and issuing the converts.
Free cash flow was a $12.4 million use of cash, reflecting ongoing restructuring and integration activities, but sequentially improved $6.5 million. Adjusted EBITDA was $28.1 million compared to $39 million in the second quarter.
Now turning to fourth-quarter guidance, we currently expect net sales of $190 million, plus or minus $10 million. Our industrial end market is expected to be flat to slightly down. The high end consumer end market is expected to be seasonally down and the infrastructure end market is expected to be flat to slightly down.
While end customer demand consisting of direct shipments plus POS has improved, our net sales guidance reflects a thoughtful reduction in channel inventories. Based on expected product mix, gross margin is expected to be 48%, plus or minus 100 basis points. Operating expenses are expected to decrease 10% at the midpoint to $74 million, plus or minus $2 million, primarily reflective of cost reduction actions we already implemented in the third quarter.
SG&A is expected to sequentially decline 13% to $33.5 million at the midpoint. Research and development expense is expected to sequentially decline 8% to $40.5 million at the midpoint. And we do not believe we have materially impacted projects that result in near-term revenue or materially affect any of our customers' roadmaps.
Net interest expense is expected to be $21 million, reflective of timing of interest rate reset periods relative to quarter end. These amounts are expected to result in a diluted loss per share of $0.05, plus or minus $0.06.
I'd now like to turn the call back over to the operator for Q&A.
Question and Answer Session
Operator
Thank you. We will now be conducting a question-and-answer session. (Operator Instructions)
Quinn Bolton, Needham & Company.
Quinn Bolton
Hey, guys, thanks for taking my question. I guess the first question, Paul, you've talked about the growing pipeline on the module side for North American operators.
Just wondering, can you give us a sense of how big is that total TAM? How much of the North American market may be currently supplied by the Chinese module vendors? Just trying to get a sense of how big this opportunity is. If you size it last quarter at $200 million, it sounds like it's bigger than that now and then I got a follow-up for Mark. Thanks.
Paul Pickle
Yeah, thank you, Quinn. By the way, I would think you get the best title on -- love the Harry Potter reference on your note this past quarter.
Okay. So getting to your question on the pipeline, I would put that market worldwide just over $6 billion. If you want to kind of strip out APAC and automotive in particular, you're talking about just north of $3 billion, I'd be excluding NB-IoT in that as well.
So if you look at the North American portion of that, there's a significant automotive number in North America and another analyst put quick tell at about -- about $4 billion over five years in terms of revenue into North America. So excluding automotive, we look at it about a $400 million opportunity.
Quinn Bolton
On an annual basis?
Paul Pickle
Yes.
Quinn Bolton
Got it. Perfect. And then, Mark, you guided the gross margin to 48%, down from over 51%. I'm just -- I guess, having a little bit of difficulty kind of reconciling that quarter-to-quarter change. It sounds like there were potentially some benefits in the fiscal third quarter that may not repeat, but it also sounds like there were some inventory charges that you took in the third quarter.
So can you walk us through what are the major puts and takes going into the fourth quarter from either the increased inventory reserves , mix among the product buckets and whatever else might account for the change quarter to quarter? Thank you.
Mark Lin
Yeah. Thanks, Quinn. I'll also point out that our guide for Q3 was a 40% gross margin. So we're guiding the equivalent margin. So there's quite a bit of noise going kind of up and down and we're quite a bit mix sensitive. So if you take the mix, normalized mix at 40%, that translates to 40% target for the next quarter.
Quinn Bolton
So really 300-plus points of benefit in Q3 that you don't expect to repeat in Q4?
Mark Lin
That's right.
Quinn Bolton
Okay. Got it. Thank you.
Operator
Tore Svanberg, Stifel.
Tore Svanberg
Yes. Thank you, Paul and Mark. So it sounds like POS is starting to improve just broadly speaking, but it also sounds like there's still some channel inventory out there.
So can you maybe quantify that for us, you know, either in weeks or in percentages or anything like that? Because it does sound like POS is a really encouraging data at this point.
Paul Pickle
Yes. I think as you reflect in my comments from the last earnings call, we talked about an improvement -- an improving bookings trend over the last four quarters. I'd say at this point, we're five quarters in.
If we look at just the IC portion of the business and the component portion, so hardware, we obviously took a significant decline this past quarter. So it's a little hard to judge where that's going to settle out. So I'll separate the two for the time being.
If we look at it, there's -- in some cases, the channels quite lean. We're in. We're well positioned to respond to any kind of market demand or market upside. In other cases, consumer, we've seen that come back. The inventory pools have been quite good. So the channels are relatively healthy.
It's hard for me to put a number on it overall because it really does depend on the product line. But if I were to speak philosophically, we would want roughly 8 to 12 weeks in the channel. In some cases, we're right at that 12-week number. In some cases, we're over that number and we're really kind of watching POS.
Lastly, on the component portion, I would say that for the first time, we've kind of crossed that one to one book-to-bill threshold. So that's very encouraging for us in terms of where we sit in the recovery.
On the hardware side, I have less visibility on the module business that's seen a significant decline this past quarter. And the reason for that is it's largely a direct business and those customers don't report the inventory number back to us. We obviously do talk to them and we do hear what they're saying in terms of what their inventory levels are.
But we've seen some decline in the industrial end market, similar to what we've seen in routers, roughly 30%. We think our customers are experiencing the same. As a result, the inventory levels in terms of weeks on demand -- or weeks on hand, have kind of gone up. And so I think it's still a little early to kind of look at what we're -- what we might characterize as a correction there, but we would expect that to play out over the next quarter or two.
Tore Svanberg
No, that's helpful color. And as my follow-up on the infrastructure business, I appreciate the guidance. They're flat to down. Could you maybe talk about some of the subsegments there and what you're expecting? Especially data center, is that expected to still grow and perhaps that's being offset by still weak PON and wireless?
Paul Pickle
Yes. So the data center definitely, we're still expecting some positive things. I will highlight that we're fairly early on, though with no change in data center. So the traction has been quite good. The channel has been pulling those orders in anticipation of usage and most of the stronger production ramps are really kind of slated for Q1 calendar.
So while we're seeing a nice uptick last quarter, we saw 112% increase in terms of data center numbers. We saw an increase this quarter in terms of data center numbers. I would expect a little bit of absorbing the pools and then kind of a resumption of growth as we kind of look at next calendar year.
So the data center number is going quite well. PON is a little bit steady state until we get a look at those tenders. We're expecting those tenders to come in in the January timeframe. That's about a one-month delay still ahead of Chinese New Year. But we were expecting kind of see those in December. It's now been moved to January. So we'll see those tenders and we'll be able to respond accordingly.
But overall, I'd say infrastructure is fairly well positioned. There are some bright spots. Most of it is moving sideways. But those increases that we're seeing definitely offset as we kind of look at the module business, although I don't expect significant further decline. We just expect that to move sideways at this point.
Tore Svanberg
Sounds good, and congrats on the progress. Thank you.
Paul Pickle
Thank you.
Operator
Craig Ellis, B. Riley Securities.
Craig Ellis
Yes, thanks for taking the question. And, Mark, thanks for the additional transparency with the convert dilution things you've posted on the website.
So, Paul, I wanted to start off with one for you and follow up a little bit on Tore's question, but maybe take a longer-term view. Would be helpful if you could provide some more color on things that you see as we look out over the course of calendar '24. In data center, for example, what are some of the things that you're really excited about, in traditional data center versus hyperscale? And then similarly, what are some of the things you have your eyes on in base station and then in PON?
Paul Pickle
Okay. Yeah. Thank you for that question. And I will say the data center, we're pretty excited. I'm personally pretty excited about data center. What we're seeing there is just an absolutely rapid SAM expansion, report expansion in data center. And I would really kind of bifurcate data center or hyperscale data centers into a general compute bucket and an AI bucket. Those are two very different applications. They have different needs.
And right now, the appetite in AI is pretty insatiable. So we're making rounds in general compute and 4 by 100 applications as well as you kind of look at just breakneck speeds in AI. And so most of the volume that we shipped today really is more 50-gig single Lambda. I know we kind of threw out 400 gig, 800 gig, but we kind of look at for us, it's a generational device thing.
We look at 50-gig devices. We're shipping 100-gig devices and we've got out in the market but are not shipping against those devices quite yet. Today, 200-gig single Lambda that's performing quite well in POCs. So as I kind of look out over the next several years, the massive growth that we expect to happen in AI, AI-driven data centers is just a significant upside opportunity for us, and we're well positioned to take advantage of that.
So to me, this is across the board, small customer set. Very exciting opportunity to say the least.
Craig Ellis
Got it. And then you talked a little bit about needing to see tenders in PON. But on the base station side of the business, anything standing out as you look at 2024?
Paul Pickle
Yes. If you go back to ECOC, we did we did a 1.6-terabit test case with the coherent demonstration and this was for 10,000, let's call it, mid-haul-type applications. So in terms of pushing up the speed or the availability to implement high-speed connections in both short reach as well as mid reach or long reach applications, both of them, we're well positioned to take advantage of.
So as infrastructure just drives data connectivity up and there's that last mile that will happen naturally as well in both European and North American countries as well, China is important to us or has been with all the infrastructure growth and multitenant dwelling growth. Their data connectivity has been really key. We would expect share in North America and Europe to creep up as they look to implement higher-speed applications as well.
So just release the 50-gig part as well. We're looking at 25-gig adoption. But I think from my standpoint, the tenders in January are going to give us a bit of a near-term effect as we continue to gain market share in other territories, which will give us a little bit more of a long-term tail as well.
Craig Ellis
Got it. And then I'll flip it over to Mark for the second topic, and it's on operating expense. So, Mark, very impressive guide decreasing OpEx by 10% quarter on quarter to $75 million. And it sounds from the color that you provided that that decrease is structural.
But I wanted to confirm that and then if you could just any color on things we should expect as we look to 2024 with OpEx, for example, fight the impacts early in the year. Any other optimizations that you might see coming? Thank you so much.
Mark Lin
Sure. Thanks for the question, Craig. So we've got at $74 million for next quarter and we're pretty confident in hitting that number because the actions that we need to hit that number have already been taken. Of course, there's always seasonality in the calendar Q1 with things like cycle resets, but we're pretty comfortable staying around this neighborhood in terms of OpEx.
There may be some incremental increases, slight incremental increases in R&D, but that's just really timing.
Craig Ellis
Got it. Guys, congratulations on the progress. Thank you.
Paul Pickle
Thank you.
Mark Lin
Thanks.
Operator
Anthony Stoss, Craig-Hallum.
Anthony Stoss
Hey, guys. Paul, I wanted to follow-up on the whole cellular module side of the business in your commentary. I'm just curious a lot of these design wins that you're lapping up now. When did it to turn to production?
And I'm curious if any of those customers are contemplating or having conversations with you about rip and replace existing products that they already have with Quectel? Or is it just kind of go-forward products? And I have a follow-up.
Paul Pickle
Yes. And so as we kind of look at it, most of the programs today that we're bidding on or have captured, they're really kind of talking about production starts near the end of the calendar year. So I would not characterize those as rip and replace.
However, if there's -- I will note that on the customer side, there's a little bit of hesitancy in terms of forecasting just because of some market uncertainty. So if there's a little bit of bolstering and market confidence, I think people start to march some of these a little bit early.
But right now, they're kind of slated at the end of the calendar year and there are additional wins and pickups that were noting, not additional follow-on programs that we've classically held. If we get some additional scrutiny out of the China Select Committee, it's highly possible that we'll see some of those critical infrastructure applications look for an acceleration of next gen or that next design.
And you know, that I would characterize a little bit more in terms of rip and replace. Because I don't think somebody will pull module out of existing hardware without refreshing that hardware as well, at least at this point.
Anthony Stoss
Got it. And then just a broader question on visibility and also if you think your January revenue guide down a little bit sequentially, you think that marks the low on a quarterly basis going forward?
Paul Pickle
Yes. So I think I am remembering my comments from last earnings call. I was pretty confident on the last earnings call that we hit a bottom on the IC business. We continue to see that we're making progress and moving forward on the IC business.
I think even though some of our peers maybe or not seeing that, we corrected a little bit earlier in the cycle and we continue to see that strength. At this point, I don't anticipate the hardware business getting any worse. I can say that.
You know, the last quarter was a little early. We're just starting to see that decline. We've seen some pretty good, let's say, steady numbers out of the routers business. The modules business did pull back as we saw a little bit of a knee-jerk reaction out of the customer base where a lot of that direct business has had.
And the Fed buying, I think was a little bit of a surprise as well for some of that customer base where our product goes into. And we've seen a little bit of sluggishness coming into the Fed buying season. And so I don't anticipate it getting worse. If anything, we might get a little bit of pickup sooner rather than later. But I think it's a little bit early to tell.
Anthony Stoss
Very good. Best of luck, guys.
Paul Pickle
Thank you.
Mark Lin
Thank you.
Operator
Christopher Rolland, Susquehanna.
Christopher Rolland
Hey, guys, thanks for the question. I wanted to talk on AI as well, just given the attention it's all getting. So last quarter, you guys mentioned the large US hyperscaler, I think, who would place some initial orders for some CDRs and TIAs for IOCs. Lot of TLAs there, sorry. But would love an update there on any purchasing.
And then you talked about sampling at some 1.60 products. Were those -- was that linear direct drive analog PAM4 CDRs? And just maybe talk about your 801.6 opportunities you see it?
Paul Pickle
Yeah, I think -- so last quarter, we talked about initial orders, but we actually saw some shipments for that large US-based hyperscalers. So that is 400-gig application. And I'll be a little bit cautious in terms of their ramp. The first half is going to be a bit sluggish, but we expect it to pick up as it goes through the end of the year.
December was final qualification hardware and then it sets the forecast. The forecast gets set according to that. So right now, things are moving rather nicely for that. That's, let's say, roughly a $30 million annualized upside once it gets the full production. And we would expect that sales cycle to span three to five years or so. So things are moving rather nicely there.
In terms of looking at -- we will see occasionally -- we do see 4 by 100s. We do see occasionally 8 by 100s. But really the next node in AI that we're talking about is 200-gig single Lambda devices. That's where you start to see 1.6 terabits in any appreciable volume.
We're looking at those speeds at both a mid-reach, you know, 10,000-type applications as well as short reach. But right now, it's just a really frothy opportunity environment for me to put a number on when those things start to translate to volume. I think it's a little bit early, but we're seeing some really some nice numbers coming out of the POCs and some demonstrations.
So given the environment, I think it's going to be -- I would expect us to be pretty quickly picked up. But I see that more of a second half calendar year '24 phenomenon rather than something in the short term. But we're going to see a steady uptick of the 400-gig modules, certainly in the first half of the year, and then it will strengthen in the second half. And then that's when we'll start to see some really -- some nice plans around 800 gig.
Christopher Rolland
Excellent. Good progress there. I wanted to look at your segments also another way here. If I'm reading kind of the tea leaves right here, on the signal integrity side, you'd expect that to fall and then protection sensing may be up. And the other two IoT-related down, is that kind of how we think about it?
Paul Pickle
Yeah. I think if we look at the APS, we saw some strong pulls, a little bit of digestion into that ramp for that smartphone -- North American smartphone manufacturer. I think we'll see a little bit of pullback, but largely due to some seasonality. That's really strong pulls up until the launch and then a little bit of pullback. So that's protection.
Protection should have a baseline component that's going to steadily improve from here. So we've seen some nice stability in the base numbers, especially as it relates to the proximity detection. And so we'll see that slowly strengthen.
I look at a 13-week booking average by product line and it's just been steadily improving, not really snapping back, but just steady improvements as those inventories come off and our channel partners can anticipate some of those production runs.
So SIP is a little bit of digestion. Signal integrity is a little bit of digestion on a really strong data center number. We could probably see some surprises to the upside. But I would -- right now, we're anticipating our revenue being a little bit softer on that channel inventory digestion.
For the initial launch, they usually like to validate the hardware in the numbers that they're seeing. And then they kind of resume an additional schedule. So protection and SIP, both kind of moving a little bit sideways at this point, but slightly down.
Christopher Rolland
Thanks so much, Paul. Appreciate it.
Operator
Rick Schafer, Oppenheimer.
Rick Schafer
Thanks, Paul and Mark. I guess my first question, if I could just maybe a high level. And, Paul, just hoping that you could share a little more about your strategic kind of vision for Semtech and what the company you think it's going to look like or you expect it to look like, say, in three years from now? I mean, what role do modules in particular play long-term and near-term growth opportunities there?
I mean, do they justify the risks with the associated margin pressure? Or do you see a path to maybe improving those margins demonstrably sort of more to kind of classic Semtech being the corporate average?
Paul Pickle
That's a good question. And you're putting me on the spot, so I certainly appreciate that. And so what do we look like three years from now? There's no doubt that our core competency really does come from components. And I do think that it makes sense to move in the module products as long as you can pick up additional stacking margins.
So to have -- to produce a module that essentially has component content that you produce does make sense. I participated in businesses before where a lot of times you have to produce a reference design. In some cases, it just makes sense to go off and sell that reference design, commercialize it, manufacture, and sell it.
You know, this gets a little bit more complicated. That's generally speaking from a semiconductor standpoint. Gets a little bit more complicated when you start talking about IoT because of the diversity of use cases, you end up having to pull a lot -- a broad range of technologies through in order to produce an entire solution. If I had to be a module maker in and of itself today, you know, the business model would be very different. And I think my investor base would be very, very different.
So does it justify, in and of itself, the margin pressure? I think the answer to that is no. But if we can pull together a bit more of a comprehensive IoT strategy in which it becomes an enabling component on higher margin sales. Then yes, I think it's quite possible that that can happen.
It doesn't necessarily mean that we have to own it for the entirety of that three-year, five-year play. But cellular backhaul is always going to be a part of an IoT strategy. And I do think that we have a rather large IoT opportunity in front of us that we need to kind of reimagine that strategy and how we're going to capture that in the marketplace.
So if we can combine it and it becomes a part of the story, along with software, along with components, then I think it makes -- it's perfectly fine, makes sense. In and of itself, no, it doesn't justify the margin pressure. In the meantime, we would look to continue to capture the opportunity. I do think that there's tremendous upside opportunity associated with modules.
In some cases, when you're engaging with utility companies, they'll want an LPWA solution or they might gravitate towards a private network like LoRa solutions. So having the opportunity to do both is an advantage as well, as long as you build together a bit more of a broader product offering or solution in store. So hopefully that gives you a little bit of a color without specifically nailing everything down after five months.
Rick Schafer
No, I appreciate that. There's a lot of color, Paul, and it actually is a great segue to my next question, which is LoRa. If you could -- if you wouldn't mind maybe give a little bit more on the order trend side, sort of order velocities, however you want to describe it.
And I'm curious, it sounded like that business was starting to pick up. I believe it's -- last call, I think you talked about it being stable sort of around $100 million or so a year. I'm curious if that's still kind of where you see that business for the foreseeable.
And if you could talk at all about sort of your strategy, if you feel comfortable talking about a strategy to grow that business, maybe even if it's a revenue funnel, if you're sharing those kind of metrics. Thanks.
Paul Pickle
So you're talking about LoRa in particular?
Rick Schafer
That was a LoRa specific question, yes.
Paul Pickle
Yes. So I kind of build as about $100 million business. So thank you for that. And so I do think as we kind of look, there's a significant opportunity for components or semiconductor content in IoT. Because of the diversity of use cases, it really becomes a multipronged strategy.
You do have to have a software component. This is where we do have a rather mature cloud platform. We do have a rather mature, let's call it, gateway software team as well. And so as we kind of look at that, there's an overall theme that I do think that there's an opportunity for us to capture, but we need to adopt this tenant of making private network deployment a lot easier.
If we look at connected in-node, connected gateway numbers, those increased sequentially this past quarter. It further shows that there is a rather large opportunity and continued wave. Even through a downturn, we're still seeing numbers from an end market standpoint that continue to migrate up where people are wanting an alternative of connectivity for, let's say, low bandwidth, low-power devices. And so that is still an opportunity that we believe in.
However, I believe I've said this before, that we need to fill out the technology portfolio. Part of the reorganization that we're going through is to kind of group R&D competencies for us to take advantage of that. And so we're in the middle of formulating that strategy in order to capture that vision of what we see, making private network deployment very easy.
And part of that story is certainly LoRa along with a few other components in the technology portfolio. We're formulating that today and we'll be working on that over the next couple of quarters, reorganizing around it. And then at the appropriate time, we'll be laying it out so that we can kind of talk about what that market opportunity looks like exactly, why we have a right to win that market opportunity, and what that timeline looks like.
In terms of overall opportunity, I know that it's been said that it's roughly we had a $1.1 billion pipeline in the past. You know, rather than qualifying that number, I'd say roughly today we're around $400 million on that lower base pipeline.
And we continue to work it. We continue to see the opportunity and the use cases. We need to look for use cases that have a lot of commonality in solution -- product solution. And then we need to start to work on delivering a standard solution to that end market rather than just simply a component.
Rick Schafer
Thanks a lot for the color, Paul. Good luck.
Operator
Tristan Gerra, Robert W. Baird.
Tristan Gerra
Hi, good afternoon. Just as a follow up to your commentary on Sierra Wireless. Just looking it up where we could see the revenue run rate before any type of synergies, including with LoRa, it looks your quarterly run rate right now is probably a bit below what it was two years ago before the big ramp-up of last year.
And you've talked about the October quarter being present would be a trough. So is a $100 million a quarter kind of a good base from which we should be basing assumptions before you build any cross-selling opportunities? And then if you could elaborate a little bit on the share shift within the different businesses for Sierra Wireless. Thank you.
Paul Pickle
Yes, I'll comment on the share shift. I don't think we've noted any share loss or share shift. If anything, we would expect to gain some share in modules, especially. If we look at routers, that's pretty steady.
The one piece, if you want to call Sierra Wireless a $400 million business, I said on the last earnings call, I believe after it recovered, I would put it right around $460 million. I still think that's probably a good number. Although, in the short term, I think your number's probably a little bit better.
So let's take the $400 million. Basically $100 million of it is going to be a pretty steady, very predictable. It's a connectivity business. It's recurring. Great visibility. We're growing it. It's not going to be a barn burner in terms of growth rate. But it's just going to be steady pickup, good gross margin, and contribution.
So the routers right now, we've seen about a 30% pullback. I'd say that's largely industry wide. I would expect that recovery to bounce back a little bit from that base, but certainly not be what it was the previous year. If you go back a little over a year, those numbers were a little bit elevated in a frothy buying season.
And the modules, that has fluctuated quite a bit. We probably put it somewhere around $260 million or so on a normalized basis and with some potential upside. We look at it. But a $390 million, $370 million upside opportunity in North America, but we don't expect to get 100% of that business. So we'll be pursuing that rather aggressively. It will take some time for us to realize that, and we would probably look to deliver those in two years out.
Tristan Gerra
Great. Very useful. And then from a follow-up question. So you mentioned making private network adoption the key here as an opportunity for LoRa. Is that really 100% of your focus on LoRa right now? As it sounds like the bulk of that inventory correction associated with helium is behind. The feedback notably from the things conferences that there is basically much easier access to LoRa implementation than just a few years ago and with a fairly clear path kind of a one to two year ROI from customers who want to implement LoRa.
So could you expand a little bit on what you think is still holding LoRa at a revenue run rate that admittedly is lower than the prior management team was guiding for a few years ago? And any elaboration on where you see opportunities to improve the adoption?
Paul Pickle
Yes. I will say this, and I think the execution in LoRa is a pretty good, let's say, below expectations that were set by prior management. But I still think anytime we grow business from 0 to $100 million in four years in semiconductor land, that's usually pretty special. So it's a unique technology.
Obviously, it's a technology that the market is there, ready to adopt. When you start bringing together a large ecosystem set of partners that all have to work together to drive a solution, it definitely is going to slow down. But as we look at adoption, it continues to pick up. There's a little bit of a blip in there in terms of helium gateways that maybe was a contributor to the numbers being a little bit larger than what actual market adoption showed during that frothy buying season
But I still look at it as good steady adoption. Overall, the market need something that's easy to implement. If we talk to customers that are looking for something that's LoRa based, they're not necessarily in IoT. They're not necessarily somebody that has the breadth of R&D in order to implement something for womb to tomb.
So this is where we have done, taken initiatives in the past in order to make LoRa easier to adopt. We are licensing the technology out in order to further that adoption. And we'll continue to evolve and deliver to the market tools necessary to speed up that adoption. But I do think that there's a couple of holes in the delivery of that solution that we could fortify in order to speed things up.
Tristan Gerra
Great. Thank you very much.
Operator
Christopher Rolland, Susquehanna.
Christopher Rolland
Hi, guys. Just a quick follow-up. The amount of turns business you guys expect to get to a guidance in the quarter, POS turns.
Paul Pickle
Yeah. Last year, we got just over 50%. We're slightly less than that this quarter. It's not something that we'll be tracking and reporting on, but it's slightly better than last quarter.
Christopher Rolland
Thanks so much.
Operator
Harsh Kumar, Piper Sandler.
Harsh Kumar
Yeah. Hey, guys. Thanks for sliding me in. Quick question, Mark, so in the analog space and the industry, we're seeing a variety of companies. Basically all of them start to guide down. I guess looking at your results, you're faring much better.
My question to you is: why is that? Is that a function of data center offsetting a lot of your traditional analog businesses? Or is it because you started the correction earlier? Or is there something else going on that is perhaps helping you do better?
Mark Lin
Yeah. It's both. I'll have Paul -- he provided a lot of color on the end markets and for data center. But it is. We did start to guide down before our peers and that group. So I think that's another big portion of why we're guiding down from $200 million to $190 million.
Paul Pickle
Yeah. I think I said it in a set of investor meetings one time. The bane of our existence is an operations team that's really good. So when we got all the orders, our team had been able to secure the capacity and deliver on it. And I'm saying that, of course, tongue and cheek, they're phenomenal, phenomenally efficient team.
But we did -- we were able to secure that capacity and deliver it. As a result, we shipped it all and saw the correction a little bit early. And as a result, we're seeing the rebound a little bit early, but we are getting helped out with a couple of key markets where we do have beachheads.
One surprisingly enough is TBS where we continue to innovate on the next-generation interfaces. We got some of the best specs out there. And yes, you do kind of share this with competition over time. But right now, we're the only guy who can deliver and we're seeing upside as a result. And AI is just a gift. It's a phenomenal SAM expansion.
Harsh Kumar
So on the AI. stuff, can I ask you if your visibility is extremely solid? In other words, are you seeing orders just pile up for multiple quarters' visibility? Or is it a situation where, of course, you're hand to mouth, but the orders are coming in kind of spurts of a quarter?
Paul Pickle
Yeah. I wouldn't say it's the orders are piling up. The design-ins are piling up. And we anticipate the orders piling up. We did see a nice bookings increase this quarter. So it's early on in that trend. It will strengthen this next year as it rolls out.
But ECOC was a nice validation on a few fronts. I think you might have seen some of the press releases that we put out, but just had some amazing feedback that came back from that. Our 200-gig single Lambda devices, very well received, very well positioned. And the industry does have a decision on which data center architectures they're going to implement.
Every hyperscaler hesitancy, will be a little bit differently. But the bottom line is it really doesn't matter what direction they go. Retimed, direct, it's all going to add up to additional SAM and growth for us. Even -- it could be a bit more explosive if they choose a particular route.
But overall, it's just a very nice frothy environment. And so we would start -- be starting to just the discussions in terms of capacity -- securing the capacity for this next year. That would be happening today and then orders would follow.
Harsh Kumar
And if I can ask one last one, do you think, Paul, that you are done cutting at this point in time? You reached a pretty happy threshold, if you will?
Paul Pickle
I'd say -- my goal was -- I intimated that we probably had another $40 million to go. We did do that. At this point, we pulled out $140 million out of the business, and that was done thoughtfully with historical budgeting exercise.
So if you go back in time, you look at 2017, you look at what the budgets were, we're at those appropriate stages. And I think everything kind of comes down to just continual refinement and improvement at this juncture. I'm comfortable with the cuts that we made. Anything that we do will just be to refine and improve our chances for growth in the future.
Harsh Kumar
Thank you, Paul. Thanks, Mark.
Paul Pickle
Thank you.
Operator
Cody Acree, Benchmark.
Cody Acree
Yeah. Thank you guys for taking my question. The next one, I guess just a follow-up to Harsh's question, Paul, what other segments you said you're reviewing today that you believe are near the bubble to require less investment going forward?
Paul Pickle
So I'd say that -- I've got a practice or cadence in my management that I have always implemented. I review everything on an annualized basis. So if we don't have good projects to invest in, we don't just assign budgeting to a business unit just because they had it before.
So every project requires a justification. Every project would have an opportunity attached to it. So at this point, I scrutinize everything on a regular basis. Obviously, I can look at market expansion in certain areas and certainly feel a lot more comfortable of the risk profile associated with that investment.
But we would look to -- as we kind of pull back OpEx overall, we'd look to refortifying the beachheads. If those beachheads have, let's say, some sunsetting attached to them, we would look to take those investments and put them in other areas that are a bit stiffer top-line growth.
Cody Acree
Thank you for that. And I guess secondly, just if you can summarize the improving demand that you've noted in the press release about your high end consumer business and data center, how much of that is for near-term visibility for the current quarter in the next six months? And to what detail can you provide that's offsetting those two businesses? I think your guidance for both was flat to sequentially down.
Paul Pickle
Yeah, yeah, that is correct some. I'd say data center came in a bit higher in Q3 than we fully anticipated. When I speak to improving demand, I'm really not talking about revenue, I'm talking about end market consumption. So that would be direct shipments plus POS. And so you just have to kind of qualify that.
To me, that's the best measure of where our business is going and it eliminates all the channel inventory noise. So when I say we have an improving demand, it's modest, sequential, improving demand that has been, for the most part, improving over the last four quarters.
That, I'd look to see corresponding booking rates -- that 13-week trailing average booking rates that match that. And I'm seeing that correlation. So anything that I give you is not forecasted. It's in the rears; it's in the rearview mirror. It's demand -- continually, sequentially improving demand that we've seen over the last four quarters, at least.
Cody Acree
All right. Thank you very much.
Operator
Scott Searle, ROTH MKM.
Scott Searle
Good afternoon. Thanks for taking my questions and sneaking me in. Nice job on the restructuring efforts.
Paul Pickle
Thank you, Scott.
Scott Searle
Paul, maybe go back to modules, it sounds like you don't necessarily need any sort of a formal exclusion list on Quectel to win business because it seems like customers are diversifying away from that anyway. But I'm wondering, I don't think I heard any comments on that front. Is there an update on that front?
And it sounds like based on the normalization of that business at around $260 million or so, that you're basically running at less than half that rate right now. When would you expect normalization on that front? I think last quarter, you talked about the middle of calendar '24. Is that what we should still be thinking about? So in the second half of '24, we're getting back to levels like that before you start to add on some incremental Quectel business?
Paul Pickle
Yeah. I was trying to give you color in terms of what the calendar year is going to do for us. So if you do the math on that $260 million and kind of tipping my hand, when I think that this is going to recover and I'd put it in the second half. So this is after some feedback from customers, what we expect, as well as some additional uptake from new design-ins.
So on the exclusion list front, no, we don't need Quectel or FiberCom to be officially excluded them in order to benefit from this. We are benefiting from it; we will continue to benefit from it. I think it has -- even if the China Select Committee came out tomorrow and said, okay, everything's fine. We're not going to put them on the list. We're still going to benefit from this.
There has been enough shift because of the overall tensions that exist with China and the US, especially around networking equipment and it's been going on for years. So it's just now finally coming to a head where people are not willing to jeopardize their company's business on the basis of using a particular supplier that might be $0.10 cheaper. It just doesn't make any sense.
So we're getting a lot of pickup. Not to mention, we have good relationships with Qualcomm. I come with those relationships because of the previous work done in compute and so we're leveraging that in order to pick up very heavily on those leads. And I should mention that we also sell quite a few Sony chipsets with the LPWA modules as well and have a fantastic relationship there.
Scott Searle
Great. And lastly, maybe to just wrap up on the LoRa front. Thanks for calibrating. It sounds like the annualized run rate now was down at about $100 million. I'm wondering what the growth rate on that business looks like and what the composition of that is today in terms of sales outside of China. I know that's where a lot of design activity has been going. But I'm wondering how that's building at this point in time.
And lastly, just on the asset sale front, I know you guys are reluctant to do anything essentially under pressure on the balance sheet. But now that you solve the covenant issues for the immediate future, are there some assets now that you're starting to reevaluate that may not be part of the long-term Semtech? Thanks.
Paul Pickle
Okay. Yeah. On LoRa, I will say we will probably still report very strong China-based number just because of where manufacturing typically is. But if we look -- if I look at that -- where that revenue originates, we're seeing quite a bit of adoption outside of APAC in particular.
And so what would the growth rate be, I definitely expect it to be on the order of 10% to 15%. In the near term, we could see some acceleration of that as a couple of technology pieces come together and it becomes a little bit easier to implement. But we've seen steady adoption in infrastructure applications, especially outside China.
We do have some China-based wins as well, but Europe definitely seeing some nice adoption there as well. So 10% to 15%, I think, is the long term, our long-term growth rate that we would probably target for that, and we can adjust that as time goes on.
In terms of balance sheet, we are definitely hyper-focused on getting the leverage ratio down. We will continue to be. I won't comment on any particular asset sale, but I'll just say that all cards are on the table in terms of driving that leverage ratio down to the benefit of our shareholders.
Scott Searle
Great. Thank you.
Operator
There are no further questions at this time. I'd like to turn the floor back over to management for any closing comments.
Paul Pickle
Thank you for joining and have a good day.
Operator
And that concludes today's teleconference. You may disconnect your lines at this time. Thank you for your participation.
