MACOM Technology Solutions Holdings, Inc. (MTSI) Q3 2025 Earnings Report, Transcript and Summary

Welcome to MACOM's Third Fiscal Quarter 2025 Conference Call. This call is being recorded today, Thursday, August 7, 2025. [Operator Instructions] I will now turn the call over to Mr. Steve Ferranti, MACOM's Vice President of Corporate Development and Investor Relations. Mr. Ferranti, please go ahead.

Thank you, Olivia. Good morning, and welcome to our call today to discuss MACOM's financial results for the third fiscal quarter of 2025. I would like to remind everyone that our discussion today will include forward-looking statements which are subject to certain risks and uncertainties as defined in the safe harbor for forward-looking statements contained in the Private Securities Litigation Reform Act of 1995. Actual results may differ materially from those discussed today. For a more detailed discussion of the risks and uncertainties that could result in those differences, we refer you to MACOM's filings with the SEC. Management's statements during this call will also include discussion of certain adjusted non-GAAP financial information. A reconciliation of GAAP to adjusted non-GAAP results is provided in the company's press release and related Form 8-K, which was filed with the SEC today. With that, I'll turn over the call to Steve Daly, President and CEO of MACOM.

Thank you, and good morning. I will begin today's call with a general company update. After that, Jack Kober, our Chief Financial Officer, will review our Q3 results for fiscal year 2025. When Jack is finished, I will provide revenue and earnings guidance for the fourth quarter, and then we will be happy to take some questions. Revenue for the third quarter of fiscal 2025 was $252.1 million and adjusted EPS was $0.90 per diluted share. We had a strong quarter of cash generation and ended the quarter with approximately $735 million in cash and short-term investments on our balance sheet. Overall, our team did an excellent job in meeting our business and financial objectives this quarter. Revenues by end market were as follows: Industrial & Defense was $108.2 million; Data Center was $75.8 million; and Telecom was $68.1 million. I&D was up 10% sequentially. Data Center was up 5% sequentially, and Telecom was up 4% sequentially. Our I&D and Data Center quarterly revenues achieved record results. Our Q3 book-to-bill ratio was just over 1.1:1. As a result, our backlog remains at a record level. Our turns business, or orders booked and shipped within the quarter, was around 17% of total revenue. We believe our backlog growth is driven by our new products gaining market share as well as a positive secular trends across our 3 major end markets. Our ability to provide competitive and leading solutions to our customers is what drives our financial performance. As a result, we continue to focus on technology differentiation across all our product lines. Simply put, our strategy is to enable the highest power, highest frequency and highest data rate applications within our 3 core markets, using proprietary semiconductor processes, IC design techniques and package technologies. In addition, over the past 6 years, our strategy has included strengthening our RF, microwave and optical systems engineering capabilities. As a result, we are better able to engage customers early on system architecture discussions rather than offer point product solutions after the schematics and block diagrams have been developed. This approach when combined with the strength of our chip designers and manufacturing capabilities allows us to have input on the system block diagram, which can translate into larger business opportunities and more cross-selling of products from our diverse portfolio. Industrial & Defense remains strong and we continue to see opportunities in the U.S. and European markets. We support a wide range of applications, including military space electronics, MILCOM, onboard drone electronics and directed energy anti-drone defense systems. Electronic warfare has been very active within our I&D business. EW systems typically contain complex wideband MMIC semiconductors generally operating at the higher frequency bands. These precision high-frequency applications often require high levels of integration and novel engineering solutions. Examples include components for radio and radar jamming as well as optical electronics using conjunction with microwave electronics to spoof radar systems. These requirements within the defense electronics sector play into MACOM's strengths. Our industrial and multimarket product lines had modest improvements in demand during the quarter. Standard product sales are increasing across a wide range of low and medium volume applications. Telecom orders remained solid, specifically in 5G infrastructure, broadband access and metro/long haul. While our lead 5G customers expect limited growth in the global radio access network market, our strategy is to gain market share with new designs that outperform the competition. In support of this goal, over the past 18 months, we have developed our next-generation high-power GaN on silicon carbide semiconductor process at our RTP fab to support existing and future 5G applications. We call this process GaN 4. I am pleased to report products from the GaN 4 process have been sampled to several of our major customers, and we have received very positive feedback on product performance. We believe our GaN 4 process will make us more competitive in massive MIMO applications, which we are not currently addressing well. Our metro/long haul and SATCOM businesses within the telecom end market continued to perform. Expansion of high-speed data transmission and increased ground-to-satellite and satellite-to-satellite communications are driving demand for our products, some of which operate at 130 gigabaud data rates and up to 80 gigahertz in frequency. And finally, our Data Center business continues to grow, driven by the global expansion of this market. Demand remains solid for our high-performance connectivity IC portfolio supporting 800G and 1.6T deployments. In fact, we believe we will have a record 200G per lane product revenue in Q4. Additionally, we are seeing demand at the lower data rates, including 100G and 400G, supporting expansion links and more traditional data center architectures. As we look at our full year results for the data center, we expect significant growth across almost all data rates and platforms. Even our legacy 25G NRZ business is expected to grow year-over-year. We are also pleased to report our data center product revenue mix is expanding as 2 new product lines enter production. First, we recently transitioned our 200G per lane photodiodes or photodetectors, also known as PDs, into high-volume production. We have a strong market position with 200G per lane PDs as MACOM is one of the few suppliers who can offer customers both the TIA and photodetector ICs, and we can offer a chip scale stacked configuration with 4 PDs mounted on top of the TIA. Second, we have secured high-volume production orders for 100G per lane linear pluggable optics or LPO chipsets. Our LPO customer is deploying an 800G network in a median reach application. We believe this production order confirms that the market will continue to evaluate and adopt LPO architectures and roll out LPO solutions at scale. LPO is beginning to spread, which is good for MACOM. Our R&D, product development and operational execution ramping the PD and the LPO products represent a major technical accomplishment by our teams. We believe these 2 new product lines will support future revenue growth. Many investors ask us about our ACC or active copper cable business and opportunities. Internally, we refer to these products as linear equalizers. As the data rates increase, we believe passive connectivity will coexist or be replaced with active solutions. Active connectivity solutions can be organized into 3 categories: active copper cable or ACC; active optical cables or AOCs; and active electrical cables, AECs. MACOM provides linear equalizers and TIA plus driver chipset solutions for ACC and AOC configurations. We do not support AECs, which are typically retimed DSP-like solutions. We believe the trends to electrify connectivity with equalization will continue to expand inside and around the data center as well as in other applications. As an example, we see some connection protocol in the compute industry moving to higher data rates and therefore, away from passive connections to active electrical or optical solutions. An example would be PCIe 7 to connect disaggregated GPUs, CPUs and memory together using high-speed connections. Disaggregated computing enables the efficient use of compute, memory and storage resources, but it increases the need for fast low-latency connections. MACOM is addressing this high-volume application with existing InfiniBand and Ethernet devices in developing ICs with PCIe-specific protocol features for plug-and-play compatibility with existing cabling and multiple connector form factors. We have developed PCIe solutions for both single mode and multimode fiber applications and demonstrated these at recent OFC and ECOC conferences. While MACOM is focused on supporting current generation applications, we are also looking ahead at future applications. As an example, at the last OFC, we demonstrated a 300 gig per lane PAM6 driver IC. As we work on advanced ICs like this, we engage with the industry leaders so we can properly understand future system requirements. Another example of advanced work includes our active design efforts and product sampling on our 400G per lane products, which we anticipate will support revenues starting about 2 years from now. During Q3, MACOM exhibited at the International Microwave Symposium in San Francisco. This exhibition is a great venue to highlight our latest microwave millimeter wave in optical technology innovations and to introduce new product lines to our customers. At this year's show, MACOM featured 16 live technology demonstrations, showcasing our newest products for electronic countermeasures, radar and SATCOM applications. I want to highlight 3 noteworthy new products. First, a 1-kilowatt X-band pulse power amplifier module developed for electronic countermeasures and directed energy applications. This is a great example of the type of high-value add system-level offerings that I mentioned earlier. Second, a wideband front-end module or FEM, covering 20 to 18 gigahertz. This multichip transmit receive module solution is ideal for wideband phased array architectures. And third, a new product solution for RF over fiber applications MACOM's [ MAT61m ] transmit receive module offers up to 70 gigahertz of bandwidth, making it ideal for antenna remoting. The module provides a complete solution for transporting wide bandwidth signals over optical fiber within a compact form factor. I'll mention a few other noteworthy items. Our RF power business continues to perform well with an attractive mix of defense and commercial applications. Supporting this activity is our wafer fab located in RTP, North Carolina. On July 25, this fab came under MACOM's full control, almost 6 months ahead of the original schedule. We felt it was in our best interest to accelerate the transfer to remove risk given the sellers bankruptcy situation. The accelerated transfer will create a modest near-term gross margin setback, which Jack will review. However, now that we have full control of the fab, we can intensify yield enhancement efforts and optimize performance and operational metrics of the fab. We have recently executed a plan to increase fab output capacity by up to 30% with the purchase of heavily discounted fab equipment. The equipment installation and qualification will take 12 to 15 months to complete. Our lead high-volume customers are excited for this capacity to come online, and we expect that this move will lead to additional high-volume program wins starting in 2026. Our business development activities based out of MACOM's European Semiconductor Center, or MESC, continues to gain momentum in the market. We believe we are better able to penetrate the major European industrial, defense, space and telecom accounts with the design and manufacturing site in France. Our goal is to be the premier designer and manufacturer of high frequency in high-power gas and GaN IC semiconductors in Europe, and I am confident our talented team in France can make this happen. And last, our 6 business units continue to engage customers on significant programs at the IC, module and subsystem level in all 3 of our target markets. We continue to strategically expand our workforce with industry-leading talent to meet the challenge, to be first to market with best-in-class performance. Jack will now provide a more detailed review of our financial results.

Thanks, Steve. Our Q3 results are at a record revenue level with strong financial performance, continuing our steady growth in revenue, increased operating income and ongoing cash generation. we have made sustained operational improvements across the business, which have supported our strong balance sheet and cash generation. Fiscal Q3 revenue was a new quarterly record of $252.1 million, up 6.9% sequentially based on growth across all 3 of our end markets. Our overall book-to-bill ratio for Q3 was just above 1:1. Adjusted gross profit for fiscal Q3 was $145.2 million or 57.6% of revenue, slightly ahead of prior quarters. As Steve had mentioned, we are pleased to have assumed operational control of the RTP, North Carolina fab on July 25, ahead of the original scheduled December transfer date. While we anticipate that the acceleration of this transfer will result in some minor near-term gross margin dilution of approximately 60 basis points or about $1.5 million in Q4. We are excited to have eliminated the business risks of not having the important facility under our operational control. With the facility now under MACOM's control, we believe we will be able to increase the speed of improvements to the fab's production capacity and yields which will help to stabilize and improve the fab's performance as we enter fiscal year 2026. Total adjusted operating expense for our second quarter was $81.7 million, consisting of research and development expense of $55.1 million, and selling, general and administrative expenses of $26.6 million. The anticipated sequential increase in adjusted operating expenses compared to Q2 was primarily driven by higher R&D associated with employee-related costs and foundry expenses as well as higher variable compensation. As we are scaling the business, we have added new capabilities and resources, primarily within R&D functions. I would like to note that we have remained very focused on controlling our OpEx as we continue to grow our revenue. Depreciation expense for fiscal Q3 2025 was $6.9 million compared to $6.8 million in Q2 2025. Adjusted operating income in fiscal Q3 was $63.5 million, up 6.2% sequentially from $59.8 million in fiscal Q2 2025. For fiscal Q3, we had adjusted net interest income of $6.8 million, increasing $400,000 sequentially from $6.4 million in Q2. Our adjusted income tax rate in fiscal Q3 was 3% and resulted in an expense of approximately $2.1 million. We expect our adjusted income tax rate to remain at 3% for the remainder of fiscal year 2025. We are continuing to assess the impact of the U.S. government's recent legislation to understand the longer-term impact on our income tax rates and associated balances. Fiscal Q3 adjusted net income increased approximately 6.1% to $68.2 million compared to $64.3 million in fiscal Q2 2025. Adjusted earnings per fully diluted share was $0.90, utilizing a share count of 75.9 million shares compared to $0.85 of adjusted earnings per share in fiscal Q2 2025. We continue to make operational improvements within the business, which can be seen in the sequential increases in our adjusted operating income and EPS over the past 8 quarters. Now moving on to operational balance sheet and cash flow items. Our Q3 accounts receivable balance was $129.5 million, down from $131.4 million in fiscal Q2 2025. The decrease in our AR balance was driven primarily by stronger cash collections. Our days sales outstanding averaged 47 days, which was below our previous quarter at 51 days. Inventories were $215.4 million at quarter end, up sequentially from $209.3 million, largely driven by inventory increases to support existing programs and anticipated future demand across the business. Inventory turns increased to 2x from 1.9x in the preceding quarter. Fiscal Q3 cash flow from operations was approximately $60.4 million, up $21.6 million sequentially and an increase of more than $11 million over fiscal Q3 2024. The sequential increase was primarily due to increased net income combined with fluctuations in working capital. As I have noted in previous quarters, given the dynamics of our growing business, it's typical to have variations in cash flow from quarter to quarter. Our business model over the last few years has demonstrated strong cash flow from operations. We believe we are on track for our cash flow from operations to be in excess of $220 million for fiscal year 2025. Capital expenditures totaled $8.8 million in fiscal Q3, up $700,000 sequentially. As a result of our increasing demand from customers during the fourth quarter, we expect to purchase $12 million of surplus equipment at the RTP fab from the previous owner. This will allow us to expand our RTP capacity by up to 30% over the next 12 to 15 months. Including this $12 million purchase, we expect our total capital expenditures for fiscal year 2025 to be in the range of $40 million to $45 million. Next, moving on to other balance sheet items. Cash, cash equivalents and short-term investments for the third fiscal quarter were $735.2 million, up $53.7 million from Q2. Comparing our cash and short-term investments to the book value of our convertible notes, we are in a net cash position of more than $235 million as of July 4, 2025. I will highlight that we expect to pay off the $161 million of our remaining 2026 notes over the next 3 fiscal quarters as these notes become due under the terms of the original agreement. As we move into our fiscal fourth quarter, we expect revenue and profitability growth and to also maintain a strong balance sheet with ample cash to support our strategic goals. We will continue to carefully manage our discretionary and capital spending to support annualized revenue in excess of $1 billion, while further improving our operational margin, EPS and cash flow. I would like to thank the entire MACOM team for their contributions over the past quarter, including those that supported the planning and accelerated integration of the RTP fab. I look forward to us making ongoing improvements to the business as we complete the remainder of fiscal year 2025. I will now turn the discussion back over to Steve.

Thank you, Jack. MACOM expects revenue in fiscal Q4 ending October 3, 2025, to be in the range of $256 million to $264 million. Adjusted gross margin is expected to be in the range of 56% to 58%, inclusive of the near-term impact of the early RTP fab transfer. And adjusted earnings per share is expected to be between $0.91 and $0.95 based on 76.5 million fully diluted shares. We anticipate 5% sequential revenue growth in Data Center and Industrial & Defense. We expect Telecom revenues will be slightly down sequentially. These targets support record revenue and earnings for the company. And finally, I would like to take a moment to announce that Susan Ocampo will be retiring from our Board of Directors effective as of August 31, 2025. Susan, together with her late husband, John Ocampo, fundamentally transformed MACOM when they acquired the company in 2009, setting MACOM on a path of innovation and growth that continues to this day. Over her 15 years as a director, Susan has been an integral part of MACOM's journey, providing exceptional guidance through our company's most significant periods. The Ocampo's legacy is permanently woven into the fabric of our company, from our corporate values to our commitment to excellence. Susan has shared with us her plans to dedicate more time to philanthropic endeavors with a particular focus on supporting engineering students pursuing degrees in RF engineering and related technical fields. This commitment to nurturing the next generation of semiconductor talent reflects Susan and John's lifelong passion for education and our industry. We're inspired by Susan's vision to give back to the industry that has been central to her professional life, and we look forward to seeing the impact of our future endeavors in the years to come. On behalf of the entire Board and management team, I want to express our gratitude for Susan's dedication and steadfast commitment to our company's success. While we will certainly miss her guidance in the boardroom, her contributions will continue to benefit MACOM for years to come. I would now like to ask the operator to take any questions.

[Operator Instructions] Our first question coming from the line of Quinn Bolton with Needham & Company.

Congratulations on the nice results and outlook. I guess, maybe start with the RTP fab conveyance, Steve and Jack, you mentioned it a 60 basis point headwind, but you now have the opportunity to sort of get in there, increase capacity, enhance yields. Wondering if you could give us some sense when do you think those improvements could turn RTP from a margin headwind into a margin tailwind?

Thanks for the question, Quinn. And maybe I'll say a few words about the transfer and then Jack can also add some comments. So number one, we're very excited to have this fab conveyance behind us. As part of that conveyance, it included bringing over and onboarding about 180 employees. So we're now welcoming the workforce that has been working in the fab and around the fab that were not technically MACOM employees over the past 18 months. So a big hearty welcome to all those employees. Our integration teams have been working very well together over the past 18 months. We've been able to migrate all the IT systems, all of the purchasing, all of the controls that was really at the -- under the control of the Wolfspeed management team is now fully under MACOM's control. So what that means is there's no more fab operating committee, where they would be sort of Wolfspeed and MACOM management defining projects in the fab, looking at priorities. All of that goes away. And now we have full rein to do what we want, when we want. And of course, the first thing we will focus on is improving yields, efficiencies, cycle times and overall quality and performance of the fab. And as part of this, I'll just highlight that we took one of our senior executives that had been running the Lowell fab and he and his family moved down to RTP. And so we have new leadership on the ground that understands how MACOM likes to run its fab. So we're very excited to see all the great work from him and his team. So as we look forward, as we start to move these enhancement programs into financial benefit, we're modeling as we look out over the next few quarters, anywhere between 25 and 50 basis points of improvement sort of going forward after Q4. That's what our crystal ball says at the moment. And so when you do that math, you can see that we're probably a couple of quarters, maybe 3 quarters away before you start to see a tailwind and then a real contribution. The other thing I'll highlight, which is really fundamental to the execution of the fab is increasing the output capacity. And so as we said in our remarks, we have made some purchases to bring in more equipment to eliminate single-point failures, open up areas where we have pinch points and capacity. And that will really allow the fab to run more efficiently, which will also be a benefit to the overall profitability of the operation. So very excited about the transfer. Did it about 6 months before we had originally planned. We had always wanted the fab to come over as sort of gross margin neutral. Clearly, we're just a little bit behind about $1.5 million on a quarterly basis right now, and we think that's easily managed going forward. And maybe Jack can add some comments around our future plans.

Great. Thanks, Steve. And Quinn, we've had the fab under our full control for all of a little less than 2 weeks at this stage. So we're still learning and finding items that are out there. We think we've got a good plan to make those improvements from a yield perspective. We are familiar with running a fab of this size. We have similar capabilities here in Lowell and that extends to some of the suppliers that we have and making sure that we're linking up our supply base with the supply base that's down there at RTP. So we've identified a number of different opportunities that we think we can work through over the next 90 days that will help improve the margins as we go forward and work our way through fiscal year 2026. So great effort and look forward to weaving the RTP fab fully into MACOM. But a lot of work was done upfront to make sure we understood what we were working through as part of this transition. And a lot of that work had helped us to accelerate it and get it done almost 6 months ahead of schedule.

Excellent. My second question, just I wanted to come back to the LPO opportunity. Very encouraging to hear. I think your first customer starting to go into volume production. Just wondered if you could talk about the pipeline you see for LPO adoption. Are you seeing a broader number of customers sort of where you're engaged, looking to go to production, say, in fiscal '26 and beyond? Just any sort of sense of how quickly you could see broader adoption of LPO.

Yes. So if we take a step back and look at some of the work we showcased at the OFC at the end of March, beginning of April this year, it's important to highlight that our demonstration at our booth had an LPO ecosystem being showcased, which included utilizing switches from 2 different vendors, servers from 3 different vendors. And we had 12 different module manufacturers showcasing their hardware which was either running at 400 or 800 gig in both multimode and single mode solutions. And so that really -- I think it puts a stake in the ground as to where LPO was back in March. And so now as we look at where we're at today, a lot of that hardware is being qualified by end users, and we're starting to see production orders roll in. And so we're very happy about that. And so we do believe that over time, we will continue to sign up new customers for this application. And then, of course, you might ask, why is that? And I think fundamental to the LPO solution is it's a solution that eliminates the DSP, which means it's lower power, lower latency, lower cost, it's easier to use and it's ideal for short lengths. Now the challenge, of course, is the customer needs to work on getting acceptable bidder rate over all of their use cases. There has to be very clean interop between all the different hardware and you lose some of the creature features of the DSP and the application. And sometimes that is a problem for the end user. So the adoption of LPO is compelling for the reasons I stated, but it also has its challenges, which means it's not ideal for all cases, but we are seeing pull, and I would say that, that pull is increasing. As we stand here today, I would say that we have signed up 1 customer for production. We're very close to signing up a second, and we expect that as we move into 2026, there will be more business.

And our next question coming from the line of Blayne Curtis with Jefferies.

I wanted to ask on the Industrial & Defense. You had strong growth in June, and it looks like it continued in September. I'm just kind of curious as it relates to 1:1 book-to-bill, if you could just talk about Industrial & Defense trends, what's driving the strength? And a lot of people with broader industrial businesses have been kind of people have been worried about whether it's starting to tail off. Just specifically anything that you're seeing in industrial would be great as well.

Yes. Most of the growth in our Industrial & Defense category is coming from Defense. So let's be clear about that. And that -- the book-to-bill of our I&D business has been overrun for at least 6 quarters, maybe 5 quarters. So the category has been doing quite well. The Industrial category for us is a little bit of a catch-all. And we did start to see some improvement this past quarter, as I mentioned, and our general business, some of that's test and measurement related, some of it's medical-related. But generally speaking, the Industrial is a small segment for us, and our main strategic focus is taking full advantage of the expanding and growing Defense market. And so we really put all our resources on that. And so I would just say that the industrial MACOM is not a bellwether for how that market is doing. It's a small piece of our business. And it's a general category and it's slightly improving today.

Great. And then I wanted to ask and I feel bad asking the question because it's been such a good segment since the acquisition. But as we kind of look at you trying to fill this RTP fab. Just curious, down slightly in September. What's driving that? And I guess, when you look at this RF business, I think it surprised a lot of people how strong it have been and you have this great opportunity to gain some share with this capacity. Maybe just comment on September and then the kind of outlook for the RF business in general.

Right. And we're very happy that our telecom business, generally speaking, has been growing really since the beginning of or the end of our fiscal '23, and we are running near record levels. And so when you sort of compare our Q3 performance to our Q4 guide, it's, I would say, noise level differences. It's not meaningful. I think what's important to note is the secular growth of SATCOM, satellite communications, where we have a very strong position on the ground, gateways as well as on the satellites. We are seeing solid business, and we are predicting growth into next year for our 5G business year-over-year. We are starting to see some improvements in our 10G PON business over the last couple of quarters. So I wouldn't read anything into the modest sequential decline going into Q4. The fact is, on a full year basis, our Telecom business should grow over 40%. So I think we're quite happy with that.

Our next question coming from the line of Thomas O'Malley with Barclays.

Congrats to the team for reaching that $1 billion run rate in the June quarter. I know that's what you kind of started out on this adventure targeting. So congrats on that. I wanted to start off on the Data Center bucket, so continued strong growth into the back half of this year. Can you walk through the moving pieces of that bucket? Historically, I think the largest piece had been kind of TIAs and drivers in optical modules and ACC was kind of coming up the curve and now you're getting some LPO contribution. But in terms of the drivers of what's continuing that strength here into the back half of the calendar year, maybe spend some time just walking through where specifically you're seeing the strength in that bucket.

Sure. And as I said in my remarks, we are seeing broad strength this year across all of the data rates. So I just want to highlight that to begin with. And I'll also remind listeners that in 2023, we had about 6% year-over-year growth. In 2024, we had 35%, and now we're triangulating for 2025 to have about 48% year-over-year growth. So very, very strong performance. We've had some great success, primarily at the higher data rates, so 1.6T, 800 gig, 400 gig. Our business and our positions there had been quite strong. And I'll highlight -- what we're excited about right now is that we're diversifying the revenue stream. And that's why I wanted to highlight that LPO is coming in. Our PDs are coming online. We see many customers moving to the higher data rates, and they're coming to recognize that it might be more affordable to use an ACC solution or to use an LPO solution than a full DSP-based module. So there's a lot of pull there. We remain focused on analog solutions. We are not -- even though we do have a DSP in production, we are not spending our R&D dollars on DSP technology. We are spending our R&D dollars on looking forward at the next data rates. And that's why I mentioned the 300-gig and the 400-gig R&D work that we're doing. So great work by the team. Generally speaking, we're engaged not only with the pluggable module people that are in the market, but also people that are working on CPO, some people call it NPO. So we are working with companies that want to push optics onto the PCBs, onto the switchboards. And from our point of view, it's an LPO solution without the package. Now the one last thing I'll highlight is our fiscal '25 was a very good year for our linear equalizers for a 1.6T application. And our lead customer there is basically during the course of this year, they ramped up that application and they ramped it down. And so when we look into our fiscal '26, we want to make sure that we are able to add new 800 or 1.6T programs to our business, and we are working to that end. So while we do see lots of growth across the product segments, I think the year-over-year comps on ACC are still to be determined.

Helpful. And then the second one is maybe for Jack, just on RTP. So you gave some of the cadence on headwind turning into a tailwind over the next couple of quarters. But I know you guys don't want to give away your secret sauce here, but I remember kind of early days of that deal, you had talked about potentially bringing some product in-house that you had done potentially externally. Could you maybe talk about the product road map at RTP? Is it something where you can bring in compound semis or make any changes to the product that's coming out of there that also offers a next leg of growth and margin improvement? Just anything that you can offer on strategy longer term with that fab and why it's so critical for you guys.

Yes. Maybe I'll say a few words first, Tom, and then Jack can add some comments. So when we first acquired the business, we recognized that their MMIC portfolio was too small. They had a lot of great foundry business, which we wanted to continue to grow, and they had a very good position in 5G. And so when we look at how to improve the margins, obviously, looking at adjusting the product mix there, primarily weighing towards MMICs, which are generally very high-margin products. We wanted to make sure that we had the best designers in the industry to make use of the processes they have. And you've seen us take actions with hiring and adding R&D resources to our MMIC design teams, including the recent acquisition of a company called ENGIN-IC, which has design centers in Dallas and San Diego. And so we really feel great about our design talent, which will drive high-value products through that fab. When you specifically asked about sort of insourcing. And so there was a part of the RF business supporting 5G includes purchasing from external suppliers, things like capacitors and passive devices, we call them IPDs, that might be used in a matched amplifier module. And we have a very active program to in-source all of that. And we will in-source it most likely in our Lowell facility, which will also solve the other issue, which is the utilization issues that we've been talking about here at Lowell. So our strategy for improving the profitability of the products includes making use of all of the technologies at our different fabs. And so some of the insourcing that I think you're referring to is more insourcing some of the silicon and GaAs passives that are currently outsourced into one of our other fabs. Did you want to add to that, Jack?

The only thing I would add is that, as we had noted, going back over the past year or so, we saw some pretty good strength from our 5G telecom customers. I think a lot of that was a result of MACOM getting involved, and they appreciated the relationship with MACOM and themselves. So that helped to expand some of the market share we had and to grow the business. There's also the I&D piece of the business or the Defense piece of the business that goes through that fab. It is a trusted foundry similar to what we have here at Lowell. So we think that's a strategic capability that we have. And as we've mentioned, it takes a little bit longer to get things moving from a defense customer perspective, but we have made great strides over the past year or so with those customers and see that as an opportunity to help support growth. And with the fab under our control, we'll be able to expand the capacity. We talked about adding some equipment over the next 12 to 18 months. So we think we're in a good spot to help improve the overall fundamentals of that fab as we work our way through fiscal year 2016 and beyond.

And our next question coming from the line of Karl Ackerman with BNP Paribas.

I was -- a lot of good questions already within datacom and industrial and the fabs, so I want to pivot a bit to telecom. I guess, why flat growth in Telecom given ongoing strength in DCI? And similar to that, when you -- I guess, when would we expect a pickup in SATCOM from some of the breadth of design wins you previously communicated? Does that begin to pick up back in December and into 2026? Any thoughts on the SATCOM opportunity within Telecom would be great as well.

Sure. Thank you, Karl. You're correct to highlight that our DCI or metro/long-haul business is doing quite strong. We have lots of programs running in terms of very high data rate for long haul as more and more data centers are being built out. It's driving the need for hardware that is, in some instances, coherent-based solutions. So that business is doing quite well. So that's really not an issue there. I think the quarter-over-quarter dip you're seeing in Q4, I would say, is more to do with just managing the backlog, generally speaking, and working with customers to hit their dates. And again, I do have to remind you, like I mentioned earlier, that year-over-year, our Telecom business in the fourth quarter, as forecasted, is doing quite well. So the year-over-year numbers are very, very solid. So I don't think there's anything fundamentally broken there. I think it's -- that market has seen some great growth this year. And we would expect -- and we've said, and we had talked about the fact that the potential within this segment is quite large. And then your last comment about how are the various satellite programs running. I did provide an update on last quarter's call that we were finishing up the MMIC design phase. We were building engineering models. That work continues. And in the near term, we're looking towards delivering that our major customer for that 1 large order, engineering models. And then once they go through a review of that hardware, we go through what they call the CDR, Critical Design Review, we locked down the design and then we start production. The timing of that right now hasn't really changed. We're thinking production could start as early as the end of this calendar year or the beginning of next calendar year. It will really depend on the results of the work that we're doing over the next few months.

Our next question coming from the line of Tore Svanberg with Stifel.

Steve, you talked about the photodetector product with the TIA and I think you mentioned the chip-scale package. I'm just trying to understand the value proposition there. I mean, it does sound like a level of integration that perhaps some of your competitors don't have. And could you talk a little bit about what specific use cases there will be for those TIAs? And what I mean by that is that sort of intersecting 800 gig or 1.6? Or is it more broad-based than that?

Great. Thank you for the question, Tore. So our photodetector product area is world-class in our -- in my opinion, in our team's opinion. It's fundamentally rooted in epi design and also the chip design itself. The chip design is backed by submicron processing that allows us to do very close alignment of all the various layers on these photodetectors. We -- our products have industry-leading dark currents, which means the products are very stable over time, very reliable over time. Our products are what they call sell self-hermetic, which means they can -- you don't need to put them in a hermetically sealed module, so they can be open to the environment, and we use very special processing to achieve self-hermeticity, which is unique, we believe. And we have a very good control over the lens fabrication process, both on the front side in aligning the lens to the backside of the device. And so a lot of engineering has gone into these designs. MACOM has been known historically for having the most sensitive photodetectors in the industry. And historically, our customer base has been test and measurement companies, people that are making optical testers that might have a sort of a gold box optical receiver in the piece of test equipment that we manufactured. And over the last year, we have been focusing very heavily on winning high-volume sockets in the data center in 2025 and 2026, you're going to see the results of that effort. And to do that, I would just add that we have really set a very nice foundation for high-volume testing in manufacturing to meet what really looks like massive step-ups in volume for us. So we're very excited with what the work that the team is doing. It's been a group effort between our business units, our operations, our applications teams and where these parts are used. So it's a 200 gig per lane photodetector, so it's suitable for an 800-gig receiver or a 1.6T receiver. So in the case of a 1.6T receiver, it would require 8 photodetectors. So you're talking about millions and millions and millions of devices.

Yes. That's great color. And as my follow-up, you mentioned PCIe gen 7. I assume these are again PCIe over fiber products and not copper. And again, from -- to sort of trying to understand the timing of that. With this, again, try and intersect, I guess, the 2027 cycle? Or could we start to see some revenues even before that?

So I -- yes, you're right that PCIe 7 is optical-based. And we are working with some customers as well on, I would say, electrical solutions that are -- that will maybe come online before PCIe 7. PCIe 7 is a fairly high data rate, bidirectional, lots of different lanes, 32 gigabits per lane, but many, many lanes. So as I highlighted in my comments, the compute industry is really pulling companies like MACOM into their block diagrams because they need to add equalization as they move high-speed data across their computer boards. And so it's kind of a perfect ancillary solution. And again, all of this would fall under our equalizers. And yes, to be very clear, we do have equalizers for PCIe that are electrical-based, not just optical.

Our next question coming from the line of Harsh Kumar with Piper Sandler.

Congratulations, first of all, on yet another solid and fantastic results. Steve, I wanted to understand the headwind with the margins. You got it a little bit earlier than you wanted. Is that simply it that the revenue isn't quite where it needs to be? And I want to understand what you can do or what you need to do to overcome this small $1.5 million or 60 basis points of margin? Are there any functional steps that are needed to be done?

Yes. I think it's -- and maybe I'll say some comments and Jack can add on. But I think it really comes down to now taking full control of the fab and being able to fix all the things that we've not been able to fix to date. And so our operations team is doing an absolutely phenomenal job moving the needle as quickly as we can. And I'll just highlight, when we first brought the business over, this business was just in a pretty rough state financially. And so it's been quite transformative over the past 18 months, and I'll highlight that even before we close the acquisition, we effectively worked with the seller to restructure the business, so that the day it came into MACOM, it would be accretive to our earnings, and we met that challenge and met that goal. And we also said that this acquisition would effectively pay for itself in 3 years, and we think we're on track as well. But as we go forward, it's looking at critical controls. It's looking at areas where we have unacceptable yields. It's turning the material through the fab faster, which means sort of reengineering the industrial side of how material moves in the fab. I wouldn't say or point to any 1 particular thing. I think it's 20 different projects that our team has in flight to incrementally improve financial performance. But the 1 thing I think we should really focus on, which is we can double the size of this business, and that is our plan. So when we look at our position in the market, in GaN on silicon carbide with this technology set, we have tremendous upside. And so it's an anchor technology that allows us to address major customers across test and measurement, military, defense, telecommunications, where we can not only sell this anchor product, but also all of our other chips around it. So it's really helping MACOM get to the next level. And so it's block and tackling in the near term. And as we said earlier, 25 to 50 basis point improvements over time starting -- beginning of next year. Jack?

The thing that I would add is just we did not have operational control of the entire fab in terms of the workforce and how we run the equipment, when we run the equipment up until a little less than 2 weeks ago. We had made some pretty good strides with our fab operating committee with the former owner of the fab. So I think out of the gate, we were doing pretty well. Obviously, over the past 6 months, the former fab owner has had a lot going on, and there may have been a little bit of a lack of focus. I think that is what drove some of our decision to accelerate the transfer. And I think we were going to be more than willing to accept that little bit of a temporary margin step back for us to ultimately be able to accelerate some of the improvements that we're anticipating now that we have full control. So -- and once again, we're able to minimize the risk now that we can control it. So hopefully, that color helps with regard to that minor margin setback that we referred to.

No, it does. I guess, my next question is also on the fab because most of the product questions have been sort of asked. I guess, I wonder if you would be willing to share with us what level of revenues this fab is doing now. And also, we understand it to be a telecom fab. But from a technical competency, what kind of technologies can you have in this fab? And what kind of markets can it address? And then also, I guess, a multipart question. When it's all said and done and when, let's say, 2 years from now, I'm talking to you, what could the margins be for this company? Could they be accretive to your corporate gross margins before you bought this? Or would it fall somewhere in the same range? Sorry for the multipart.

Harsh, these are great questions. And before I answer or try to answer many of those questions, I'll also highlight something that we're very happy about, is as we exit our fiscal 2025, we're really running at operating margins that are peaking really in the last couple of years. So even though we had a little bit of a setback here on the gross margin, $1.5 million of COGS or 60 basis points, the business and the overall fall-through in the operating income of MACOM is growing. And as we look into next year, we believe that the earnings are going to grow faster than the revenue. So I think we have very good control over a lot of the different moving parts. When you ask about the revenue by the fab, that's probably a little sensitive to discuss in any level of detail. We did announce when we acquired the fab, what the prior 12-month run rate was, but we really haven't updated the business since then for competitive reasons publicly. When we look at the strategic value of the fab, and I think that it is not just a telecom fab. They have -- they -- when you look at their core technology, it's essentially very high-power GaN on silicon carbide that operates up through X-bands or around 10 gigahertz on most of their -- most of the legacy processes. Those are ideal for very high-power applications. A lot of that technology can replace LDMOS, especially in military applications, where you have high -- you can achieve higher powers and get better efficiencies. The RTP fab also has a 0.15 micron process for microwave applications. And that process operates up to about 20 gigahertz. So now you're able to capture backhaul radio links, you're able to capture other telecommunication applications, military applications, and it covers many SATCOM bands. SATCOM is one of our fastest-growing areas. We talked a lot about it over the last couple of years, with more and more LEO constellations being launched and MEO constellations. The RTP technology is ideal for those applications. And I'll also add that many of these platforms are now adding direct-to-sell or direct-to-device connectivity, which typically run at the cellular bands. And that is where we really have an advantage, where we have a very strong position in 5G. And many of these LEO satellites are effectively flying base stations, where they need transmitters and receivers and MACOM has great technology to support that. And then the last thing I'll add is, it is a trusted foundry with a very high MRL level for -- to support military production programs. And so that business for MACOM is booming. And it's booming because we have a very strong amplifier in transmitter design capability. And when you combine that system-level capability -- engineering capability with the GaN on silicon carbide technology at RTP, we're hard to beat. So we're very excited about the prospects. And as I said earlier, I'm confident we can double this business -- double the revenue of this business.

Our next question coming from the line of David Williams with the Benchmark Company.

Thanks for letting me jump on. I guess, maybe first, thinking about the lower speeds, you talked about that are gaining some traction. What's driving that? Is it more of just the expansion beyond the core data center and into maybe edge AI as kind of moving out or maybe anything that's driving the legacy type solutions?

Yes. I think your thinking is correct there. We believe that as the AI data centers basically begin to expand, there's sort of a front-end traditional data center maybe in front of it, looking towards the Internet. And so we are seeing an uptick across various product categories that would support that forward-looking -- the forward-looking equipment. And so we think that the lower data rates are being sort of dragged along because of the build-outs, and we believe that's why our lower data rates are growing.

Okay. Great. And I would assume that those products probably have a little better margin profile. And then just secondly on that, can you talk to -- is there -- what the impact is from the utilization or underutilization at the Lowell fab?

Yes. So -- well, 2 things. The lower data rate products are in most of our drivers and TIAs are not processed in MACOM fab, so I want to clarify that. So it's not necessarily impacting the Lowell utilization. And then your question about, are the margins higher at the lower data rates? We don't -- we probably can't comment specifically on that. And over time, as volumes go up, typically, prices come down, that's a general trend. But at the same time, with our yield enhancement programs and working with our supply chain, we're able to drive cost down. So it depends on the product, it depends on the customer and the application as to whether these lower data rate products are sort of higher or lower margin. It depends.

Our next question coming from the line of Harlan Sur with JPMorgan.

Great job on the quarterly execution. I know there's been a lot of questions on the RTP fab conveyance, but I would have thought that there would have been a gross margin step-up on the transfer. I mean, you've had a supply agreement in place. I assume that your cost per wafer under the supply agreement was low speeds fully loaded costs, right, which includes a lot of the deficiencies and inefficiencies you've talked about yields, et cetera, right? But then they add of markup. So now with full ownership, you don't have that markup, which can imply better gross margins. So what am I sort of missing here?

Yes. So our goal was always to have the fab come over as neutral or positive. And so we've clearly missed that goal, it's coming over, and we're about 60 basis points away from where we would want it to be neutral. I'm not going to comment on the supply chain agreement or the cost structures on our prior arrangements with Wolfspeed when we were not in full control of the fab. So I really can't comment there. I think it's important to highlight that, as we've said now multiple times that as we look forward, we expect to improve the performance of the fab, which will drive gross margins. And remember, MACOM is a complex company. We have a fab in Lowell, we have a fab in RTP, we have a small fab in France, and we have our fab in Michigan. And you combine that with a large business that is effectively a fabless business, there's a lot of moving parts in our model and our gross margin model. We like the diversity of our business and our manufacturing footprint. We think it's a competitive advantage. And as we review the execution of the acquisition, the integration, the transfer, we give ourselves a very high grade for getting to where we're at today, and we think things will only improve from here.

Perfect. And a lot of product questions and technology questions already asked, but foundry, has always been a part of the MACOM strategy, right? It's been a while since we caught up on this part of your business. I believe, the team offers, I think, the most diverse set of III-V compound semiconductor processes in the industry, right? I think you guys have something like 20 different process types. You support a number of different device architectures, dials, MOSFET, MESFET, pHEMT, like what's the traction been like in attracting foundry programs? How big is foundry as a part of your overall revenue profile today and maybe your view on the growth outlook for this part of your business?

Harlan, I'm very impressed that you know all our processes, so thank you for that. You're exactly right. We have a very rich portfolio of processes, and I would perhaps invite investors and listeners to look at our summer newsletter, which we put out, and it highlights a few different things. A lot of the work we're doing with EW systems, BAE award. It talks about a contract we received from the French government. And it also highlights and summarizes the various processes that we have at our various fabs, whether it's gallium arsenide, GaN, some of the AlGaAs devices that we use for our limiters in some of our I&D customers as well as some of our very high frequency GaN on silicon. So you're exactly right to highlight that there is a lot of diversity. And our philosophy regarding foundry is to allow customers access to our fabs as foundry customers. MESC has always had a foundry customer base in Europe. We're continuing to build that. Here in Lowell, we have a small number of foundry customers, mostly test and measurement and defense, and we welcome their business and we support that business. And of course, the fab in North Carolina also supports foundry. So our business model is to allow customers to design chips if that's the way they want to run their business, and we welcome their business as a supplier. We don't break out the foundry numbers specifically, just again because of competitive reasons.

Our next question coming from the line of Tim Savageaux with Northland Capital Markets.

I know we're running long here. But just a quick question on Telecom, which at least came in a little bit stronger than I expected. And I'd be curious, in particular, what you're seeing in cable networking, that does seem to be picking up a bit across the industry. And I'll just ask my follow-up real quick here. With regard to LPO, you mentioned new engagement there. And on the topic of the business diversifying, is that with a new customer or a current large customer? Can you give us a little more color there?

Thank you. Yes, I'll start by saying our cable infrastructure business is growing. It's a small piece of the overall telecom number, but it is going in the right direction, and it is growing, so we're pleased about that. The LPO engagements that we have are with customers that we previously had done business with on other platforms. So they're not foreign to MACOM. We have been supporting them with our other products for some of their pluggable optical modules that are more traditional modules. And as I mentioned in our script, we have 1 lead customer and we think that the dam is beginning to break and we'll be bringing on additional customers soon. I'm not sure I covered all of your questions, but I'll pause there.

Our next question coming from the line of Richard Shannon with Craig-Hallum Capital Group.

My first question would be a quick and simple one, a 2-parter here just to get a sense of size of revenues here. I'd love to get a sense of the split between I&D, especially the defense here, it sounds like it's going very well, would love to get that. Then also especially with the close of the RTP fab and you talked a lot about GaN today, what's the percent of sales coming from your various GaN processes and products?

Yes. With regard to the I&D split, that's varied over time. I think if you were to go back a number of years, we probably had a majority of our revenue coming through on the industrial side of the business. That's obviously changed over the past couple of years as the defense piece is picked up and also through some of the acquisitions. So we're probably closer to a 65-35 split over time.

And then on GaN, Jack?

Yes, we don't generally break those pieces out as part of our public discussions.

Okay. Fair enough. I thought I'd try that one. My follow-on question is, as people are trying to model for 2026, I'd love to get a sense of relative growth by your segments here. And maybe I'll offer a couple of thoughts here as I was trying to work through this. You're obviously seeing some tough compares in the Telecom business. In Data Center, it sounds like you're adding some new products, which could continue that growth there. And obviously, Defense is going well. But I know you're not going to give us much quantification, but if there's any relative growth by segments that you offer for next year, that would be a great outlook.

Yes. Thank you for the question. It's certainly difficult to discuss 2026 at this stage. I'll just highlight that 2025, we should be over 30% growth, maybe closer to 32% or 33% growth. All of our secular growth trends are intact across our core markets. Our portfolio has room to grow. We are still a small company relative to the $8 billion SAM that we sort of stand in front of. I would say, at a high level, our revenues are slightly ahead of maybe what we talked about previously in terms of achieving the $1 billion. We talked a lot about the gross margins being slightly behind, and we acknowledge that, but our earnings are on track. And I think as we achieve that sort of $1 billion run rate, we should have earnings of over $4 a share as we improve the things that we need to improve. Clearly, Data Center is volatile. We've talked about that in the past. That hasn't changed. Our position in the Defense market, Telecom market is very, very strong. I think we touched on some of those. Probably too early to talk about what markets will drive our growth next year. I think that's more likely a conversation we'll have on our next call as we wrap up our fiscal '25 and then look forward into 2026.

Our next question is coming from the line of William Stein with Truist Securities.

I fear I might have misheard the last one, and I hope I'm not asking the same question, but there's been a lot of attention paid today on this call, I think, to the gross margin effect of the RTP fab conveyance. We understand there's one other aspect of the business that you're preventing gross margins from achieving the -- the targets that you established at some point were 60% plus at a $1 billion run rate. And you're guiding lower than that. So part of it, I get it, it's the RTP fab conveyance. I think the other piece is utilization in Lowell. But can you talk about your ability to track towards that 60% level. You discussed improvement in utilization in Lowell and performance in RTP. So should we expect perhaps sometime in the middle of next year to hit that 60% bogey?

Yes. Thank you for the question. And so as we look out into '26 and I realize it's early to be doing that, but as we look at the programs that we have in place and all the moving parts, we think we're more likely to exit 2026, closer to 59% gross margins in that sort of a year from now. And so that is sort of the trajectory we think we're on. So I don't think the 60% gross margin is going to happen in our fiscal '26. It's more likely a fiscal '27 event.

Okay. A follow-up, if I can. You've talked a lot about all the moving parts within your data center business. There's a lot of products there. And for some of us the detail is maybe a little overwhelming. So let me ask a more high-level sort of simplified view of this. CapEx is still very strong in that end market. You're guiding to a level where we should expect something about 48% sales growth this fiscal year. If there were no great movements in products, I wonder what we should expect growth to be next year. I would assume you don't want us modeling a similar 48% growth in the coming year as well. That's quite optimistic. But can you talk about the puts and takes and how you might encourage us to think about growth in that market next year?

Yes. And I think you're right to say that we wouldn't recommend putting 48% growth into year-over-year growth for fiscal '26 related to the Data Center segment. When we look at our business, we want to have a business that's constantly improving its profitability, improving its operating margins, constantly diversifying the portfolio and getting a stronger position in the market, and data center is one element of our strategy. Big picture, we think MACOM should be growing double digits on the top line. And so you've seen that performance over the past few years -- 4, 5 years, we would expect that to continue. The mix will change every quarter and every year. But because of the things we've talked about today, including the heavy investment in the technologies and the products for our core markets, we think we have a very good strategy and a very good team that can execute the strategy. So how that shakes out for the data center, we'll have wait and see. We'll take that sort of one quarter at a time. But I think big picture, which I think you were asking about, it would be our expectation that we are growing our top line by double digits, and we're beating -- we have even a higher growth rate on our bottom line.

And I'm showing no further questions at this time. I will now turn the call back over to Mr. Daly for any closing remarks.

Thank you. In closing, I'd like to thank all of our employees for making these results possible. Have a nice day.

This concludes today's conference call. Thank you for your participation. You may now disconnect.