MACOM (MTSI) Q1 FY26 earnings call: guidance and Q&A

What gives MACOM confidence to raise its FY26 Data Center growth outlook to 35-40%? Is it market growth, visibility, or share gains?

The confidence stems from a combination of factors, primarily the strong activity and design wins in 1.6T products transitioning into production. The company has a healthy backlog, with the second half of FY26 expected to be stronger than the first, indicating robust demand and program ramps.

“But the key underlying driver is 1.6T. That's where we see the most activity, the most design wins transitioning into production runs.”

asked by Quinn Bolton · answered by Stephen Daly

How can MACOM benefit from a shift towards CPO or NPO architectures?

MACOM's product set for CPO/NPO is similar to pluggable modules (drivers, TIAs, photodetectors, lasers). The shift benefits MACOM due to increased channel complexity per chip, which aligns with their design capabilities. The company is making progress with CW lasers for silicon photonics solutions and sees opportunities in coherent modulation for these systems.

“But we do think that the CPO and the NPO sort of transition is a benefit to MACOM. I'll also add that from an overall architecture, you see many, many channels in a smaller form factor.”

asked by Quinn Bolton · answered by Stephen Daly

What are the growth prospects for the Telecom segment this year, especially considering NXP's exit from the RF power market?

The Telecom segment is expected to grow high single-digit to low double-digit in FY26, following 40% growth last year. While the overall 5G RAN market is flat, MACOM aims for share gains from NXP's exit, which will take 1-2 years to materialize. Significant growth is also expected from SATCOM, particularly LEO programs, with opportunities extending into 2027 and beyond.

“This year, we don't expect the same level of growth. It's more likely going to be high single digit, maybe low double digit, but we'll have to wait and see on the timing of some of the programs that I've talked about.”

asked by Vivek Arya · answered by Stephen Daly

Can you provide an update on the RTP fab's contribution to gross margin and the progress on in-sourcing components?

The RTP fab has consistently improved profitability since acquisition. Future gross margin improvements will primarily come from increased utilization of Massachusetts and French fabs. The RTP fab's main goal is to increase output by 30% through cycle time reduction. Benefits from in-sourcing components are expected in 2027 and beyond, as they require new product design cycles. The French fab's 3-inch to 6-inch wafer transition is expected to be fully qualified by June 2026, also contributing to future benefits.

“The improvements that we're seeing this year and going into next year for gross margins will, I think, primarily revolve around improving the utilization of our Massachusetts-based fab and to some degree, our French-based fab.”

asked by Thomas O'Malley · answered by Stephen Daly

What is the current status of MACOM's LPO business, including commercial revenues and market size potential over the next 2-3 years?

MACOM is very bullish on LPO, with three hyperscalers now in various phases of production for 800G modules (100 gig per lane). While LPO is currently a small part of the market, it is expected to remain relatively small in the next 1-2 years but grow over time, especially as it evolves to NPO and CPO. The company is also seeing interest in LRO at 1.6T using 200-gig per lane chips.

“So we are very bullish on LPO. In fact, we now have 3 hyperscalers embracing LPO, and we are in various phases of production with those 3 hyperscalers.”

asked by Harsh Kumar · answered by Stephen Daly

Is the strong Data Center growth driven by market expansion or share gains, particularly for 1.6T analog components?

The growth is primarily due to the expanding market and MACOM winning designs, rather than significant share gains against competitors. Key drivers include the 200-gig per lane portfolio and optical components like photodiodes. MACOM is also developing CW lasers for silicon photonics solutions, which would represent new market share as they don't currently sell lasers into 1.6T applications.

“I'm not sure it's so much share growth per se. I mean as we look at our dashboards and where we have content and where we don't have content, we see competitors on all sides. And so I wouldn't necessarily say that this is a share gain. I think it's -- the market is growing. We're winning designs.”

asked by Blayne Curtis · answered by Stephen Daly

Are the two customers confirming CW laser electrical requirements module makers or hyperscalers, and what is the next step?

The customers are primarily module makers. The next steps involve collecting reliability data, and once satisfied, the module makers will go through a PCN process with hyperscalers to get MACOM's lasers on their approved vendor list. This marks a significant milestone as MACOM now has a compliant laser, though it's still in the early stages of the production process.

“I mean I would say that we're engaging with more of the former. So it's really the module customers that are our first line of entry. So we want to make sure that they are able to use our laser. They're getting good module level results.”

asked by Sean O'Loughlin · answered by Stephen Daly

Can you elaborate on the growing interest in linear equalizers for PCBs, and how their economics compare to ACCs?

Linear equalizers are gaining interest for backplanes to eliminate retimers in compute connections and for 400 gig per lane speeds where equalization is crucial due to trace losses. While compelling, the volumes for backplane applications are not expected to be as large as the Active Copper Cable (ACC) opportunity, which is seen as a bigger architectural growth driver with larger potential volumes.

“So adding the linear equalizers to various backplanes is of interest to lots of customers not only, let's say, in the traditional sort of AI construct, but also closer to the -- as you get into the compute, we see opportunities to eliminate retimers in various compute connections.”

asked by Christopher Rolland · answered by Stephen Daly