Aug 25, 2026 earnings call — Data center ramp and portfolio optimization
Guided net sales of $410 million, plus or minus $5 million, up 20% sequentially.
Revenue $0.3B (est $0.3B) · EPS $0.71 (est $0.61) · Next-quarter revenue outlook $0.4B
Key quotes
"We are at the center of one of the most significant infrastructure build-outs in history, and our portfolio plays an essential role."
"The announced sale of our cellular module business is a significant step in our portfolio optimization, allowing us to more sharply focus on our core growth areas."
"LoRa-enabled net sales were $58 million, up 31% sequentially and up 58% year-over-year, another all-time record."
Analyst Q&A 10
data center capacity needs
Q (Quinn Bolton · Needham & Company): Can you support near-term data center growth and secure capacity for fiscal 2028? Also, will you need wafer prepurchases or back-end prepurchases, and how is Semtech positioned for NPO solutions?
A: They have enough near-term capacity but the capacity secured may not be enough to support FY '28, especially the second half. They are working with front-end and back-end manufacturing partners to increase capacity, adding testers and qualifying additional partners, and can use prepayment, CapEx or other means. They are engaged in about 10 to 15 NPO-related programs and are expanding into photodiode arrays and co-optimization with TIAs.
"Now with the strong booking momentum and record backlog, we see the capacity we have secured may not be enough in supporting the FY '28, especially second half of FY '28." — Hong Hou (CEO)
LoRa growth and HieFo capacity
Q (Richard Schafer · Oppenheimer & Company): Is 20% still the right growth target for LoRa given recent strength? Also, update on HieFo capacity, design funnel, and how CW laser channel density affects demand and competition for drivers and TIAs.
A: LoRa grew faster than 20% year-over-year and they expect year-over-year growth to be better than 20% and sustainable. HieFo has been owned about 5 to 6 months, the line has been upgraded, customers' confidence improved, high-power lasers have been sent to 5 or 6 module makers, SOAs are being sampled, and fab capacity will increase about 3 to 4x by the end of the year alongside adding testers in the back end.
"Then for the fab capacity, we're just fortunate to be able to acquire an already fully facilitized fab in the close proximity to the current facility." — Hong Hou (CEO)
data center product mix and margins
Q (Sean O'Loughlin · TD Cowen): What parts of the product portfolio are driving the data center strength, and how does that align with your TIA share on 1.6T? Also, how much of the gross margin expansion is mix versus structural and is it sustainable?
A: The data center strength is from multiple areas: CopperEdge redriver/equalizer, FiberEdge share gains at 800 gig and 1.6T, drivers coming, and new photonics after the HieFo acquisition. The sale of the cellular module business should result in over 500 basis points of gross margin improvement, and much of the recent quarter-to-quarter margin gain was due to mix; post-divestiture a good starting point is about 64% adjusted gross margin.
"The sale of the cellular module business, I expect will result in over 500 basis points of gross margin improvement." — Mark Lin (CFO)
what drove data center upside
Q (Christopher Rolland · Susquehanna): What specifically caused the data center upside versus prior expectations—1.6T, LPO, TIA attach, or CopperEdge? And when will HieFo products like CW lasers, photodiodes, and SOAs ship in volume?
A: The upside came mainly from an earlier inflection in 1.6T FiberEdge and faster-than-expected qualification timing across module manufacturers, which accelerated customer demand and backlog. Currently the gain chip is shipping in volume, and high-power CW lasers and SOAs will be available for sampling and qualification in a couple of months, with significant content increases expected around the 3.2T transceiver cycle.
"Right now, the ongoing product shipping in volume is a gain chip, and we're going to be having the CW high-power CW lasers and SOA available for sampling and qualification from the customer side in a couple of months." — Hong Hou (CEO)
3.2T timing and product content
Q (Harsh Kumar · BMO): When do you expect 3.2T design and deployment, and why do you describe content as high double-digit? Also, what should we expect from ACC and LPO growth?
A: High double-digit means about 80% to 90%, not 18% or 19%. 3.2T design windows start opening in about 12 months, early movers around 18 months, and meaningful deployment in about 2 years; 1.6T and 800 gig have healthy runways for the next 2 years. ACC volume deployment starts in Q4 with cable manufacturers increasing orders and board-level linear equalizer designs happening; LPO has produced meaningful revenue since Q1 and serves as a proof point that will evolve into NPO/CPO/XPO integrated forms.
"So probably I confused you, just high double-digit means 80%, 90% instead of 18%, 19%." — Hong Hou (CEO)
800 gig strength and pricing
Q (Joseph Moore · Morgan Stanley): How strong is 800 gig demand and how long will it persist? Also, any pricing changes on the optical side?
A: Industry forecasts for 800 gig rose from about 50 million to as high as 80 million or 90 million units, and Semtech has a very healthy backlog for FiberEdge to support 800 gig with customers increasing demand. Pricing is favorable with no expected erosion in the near term for orders in the backlog, and they can generally pass slight cost increases to customers while maintaining long-term partnerships.
"And we have a very healthy backlog for FiberEdge to support 800 gig." — Hong Hou (CEO)
margins of new products
Q (Tore Svanberg · Stifel): As data center mix grows, how will newer products like photodiodes and CW lasers affect margins? Will they be at or above corporate average?
A: They plan to provide more details at the October Investor Day on mix and TAM. Newer products such as CW lasers are expected to carry data center gross margins that are accretive to the corporate gross margin average.
"So all the areas that you just mentioned, CW lasers, especially, they're at a data center gross margin, which is accretive to that corporate gross margin average." — Mark Lin (CFO)
1.6T mix and customer bookings
Q (Craig Ellis · B. Riley): How should we think about 1.6T as a percent of mix now and next year? And do customer bookings reflect you becoming strategic or supply scarcity?
A: 1.6T already surpassed 50% in Q3 and is expected to grow north of 50% while 800 gig remains strong and Tri-Edge products also contribute. The acceleration in bookings reflects product performance, availability, and services that have increased customer confidence and helped Semtech gain share.
"In Q4 -- Q3, the 1.6T is already surpassing 50%." — Hong Hou (CEO)
backlog length and ACC adoption
Q (Cody Grant Acree · The Benchmark Company): How far does the backlog extend into fiscal 2028 and can you secure more capacity for upside? Also, how broadly will ACC be adopted beyond the lead hyperscaler and what challenges remain?
A: For the remaining of this fiscal year their target is fully booked, and for the next fiscal year they have probably over 70% booked with strong momentum and confidence. There is room for capacity upside between 50% and 100%, ACC adoption will broaden over time starting with leading hyperscalers, and availability and interoperability are the important near-term considerations.
"So the backlog for the remaining of this fiscal year, I would say, for our target is all booked." — Hong Hou (CEO)
LoRa mix and portfolio plans
Q (Scott Searle · ROTH Capital Partners): How has the China mix changed for LoRa and how big is Amazon in the LoRa outlook? Is the LoRa upside mainly Amazon, LoRa Plus, or other factors? Also, will router/gateway and IoT platform remain core after module divestiture?
A: Most revenue is still LoRaWAN while LoRa Plus now represents about 20% to 25% of LoRa revenue; Sidewalk is currently nominal and this year will probably be high single-digit but has large potential if deployed at Amazon consumer scale. Portfolio optimization is ongoing; after the divestiture they like the remaining portfolio but will continue to evaluate further optimization while focusing on closing the cellular module sale.
"And Sidewalk is still, at this point, is nominal and this year will probably be high single-digit, but again a lot of potential once we are able to piggyback into the consumer at Amazon scale." — Hong Hou (CEO)