Please welcome Rich Puccio, Chief Financial Officer of Analog Devices, interviewed by Harlan Sur at JP Morgan.
I am a giant. Stand up on my shoulders. Tell me what you see. I am.
All right. Good afternoon, everyone, and welcome to JP Morgan's 2026 U.S. All Stars Conference here in London. Again, my name is Harlan Sur. I am the semiconductor and semiconductor capital equipment analyst in the U.S. for the firm. Again, very pleased to have Rich Puccio, Executive Vice President and Chief Financial Officer at Analog Devices, here with us today. For those of you that do not know, the Analog Devices team, leader in high-performance, mixed-signal, RF, analog semiconductor solutions, strong position in power management, very strong position in signal chain processing, which is both analog and digital, which is a technology that bridges the real world to the physical world. Best-in-class growth, operating free cash flow margin, strong capital return program, and a very diversified business, industrial, automotive, data center, comms infrastructure, greater than 85% of the company's total revenue. Rich, thank you very much for joining us today.
Thanks for having me, Harlan.
I think the best place to start off with is where we are in the current upcycle and the resilience in the business, right? As the industry and Analog Devices emerge from a cyclical downturn in 2023, 2024, your business bottomed in April of 2024. You were driving about 34% year-over-year revenue declines, but did start to drive positive sequential growth trends, which continued all the way into this year, right? You drove the positive year-over-year inflection in the April quarter of last year. You have driven seven consecutive quarters of accelerating year-over-year trends, anticipating 41% year-over-year growth this quarter. Take us through the dynamics over the past 12 months since we last had you here, and more importantly, your view on the idiosyncratic and cyclical dynamics in your business going forward.
A lot of questions. I guess a little bit of it starts where Harlan started. As we started to exit 2024, we had started to signal and started to see some of what I would call idiosyncratic drivers that were areas in our business that were holding up better than others and starting to show growth. Early on, we started talking about our ATE business and our Aerospace & Defense business. As we looked forward, at that point, we were anticipating that we would start to see growth there, and sequentially, with continued improvement, the broader cycle upside would come into play. What happened as we progressed, two things happened. One, we did start to see that cycle improvement, but I would say it was not as early as some of our idiosyncratic pieces.
I will go back to some of the key signals and things that we were talking about as we evaluated this.
Yes
This change. If you just look at where the, we will talk about this more as we go, I am sure. The areas where we are calling it sort of the super cycle part of the business. Aerospace & Defense, obviously, you can see everything going on and around the world and the increased demand. That was very resilient, and that was growing. ATE was starting to see significant benefit from the increased infrastructure spend. This was early, sort of the indirect impact of the data center infrastructure build-out. Because of the increase in high-bandwidth memory, high-performance compute, you need more complex testers, which meant you needed more ADI content in the testers to be able to do the complicated tests. So those two businesses were really strong.
As we continued to move forward throughout the year, what we started to see was obviously the acceleration in our data center business that we were seeing as part of that build-out. We will talk more about that. That is a business where we have been a very strong player in the power and the optical side. So that started to grow. As the broader cycle markers started to show, one of the things we talked about paying attention to determine where we were in the cycle was, when did our broad market industrial come back? We are, at our core, a 50% industrial products company. For a long time, we were not seeing any of that mass market, high-volume product coming back to historical volume levels.
Well, we started to see that come back, then what you started to see was the real acceleration in our growth because we had our idiosyncratic drivers adding the third, which was now data center, and also the cyclical drivers.
Right.
We had a couple of businesses within end markets that were slower to recover, but by the time we get exiting 2025 and into 2026, we are now in growth across pretty much all of the end markets.
That is right.
and all of the sub-parts of the end markets. A good one, a big chunk of our industrial business that was slower to recover was our automation business. Of course, we started to see automation grow, and now for the last almost year, automation has become a big growth business, and that's a pretty good slug of our industrial. But if you think of, essentially 30% of our business is in ATE, Aerospace & Defense, and data center, and those have been growing at very high rates. Then to have the broader recovery, you start to see that come through. The other area for us, which we've talked a bunch about, is automotive, right? Despite what has been a relatively soft SAR units environment in auto, we've continued to outgrow units pretty significantly.
If you look over the last five years, we've outgrown units by about 15%, and that's largely because we've got share and content gains, right? We made investments early in the products that support automated driving systems and the immersive cabin experience.
Yes.
Those have been really strong growers for us. Then you start to step back and look at it. We feel very good with the demand signals we're seeing.
where we are from a booking and backlog perspective, that we've got an opportunity for sustained growth in the double digits, even into 2027.
Yeah.
Look, there is a lot of uncertainty in the macro environment, so we always have to be cautious about that, but with what we see today, I do not think we have ever been more confident in where we are positioned in the market, and looking out at the opportunity next year.
No, that is great. To your point, I think last quarter, Q2, the team drove double digits year-over-year growth in all four of your segments, right? Talking about a synchronized sort of recovery across all of your different end markets.
Yeah. Like I said, an early signal for this is we have been talking about since 2024. We continue to see really strong results from a design win perspective.
Yeah. Mm-hmm.
And if you look at the growth in 2025 in design wins and 2026, that's another good indicator for where we're headed medium term. In addition to our end markets, the other thing that has happened is we've seen growth return in all of our geographies.
Yeah.
It's very broad-based at this point, across all of our end markets and across all of our geographies.
Well, maybe more from just a near term perspective, right back last year when we hosted you at this conference, the team had already driven sequential bookings growth in six out of the last seven quarters. That's now 10 out of the past 11 quarters of sequential bookings growth, right? You're about midway through this quarter. January quarter, next quarter, typically does tend to be seasonally weaker sequentially, but do the current order trends imply a return to seasonal trends in January? How are you seeing sort of order activity, current order activity by end market or geography?
I'll go backwards. I don't do any mid-quarter updates on our bookings, and I also don't guide to Q1. However, you are 100% correct. Typically for us, Q1 is a down market. But what I will say, and I think I might have said this in my Q&A on the earnings call is, with what we see today, I do expect that we can deliver an above-seasonal Q1.
Got it.
Normally Q1 would be down in sort of the mid-single digits for us.
That's right.
I do think we can do better than that given what we've seen. One of the things we've seen and I think you probably, talking to other semis, are seeing a little bit of this, we're getting a little bit more visibility than we probably would've had six months ago, Harlan, where I'd be very confident in what my one quarter out visibility looked like, we're starting to get a bit more bookings coming in out into the second quarter, which just has given us a little bit more visibility. Which is helpful for us given we're making wafer starts six months in advance of when we're actually delivering wafers. So that visibility's helpful for us.
The team has done a good job of shipping below your customer and channel consumption levels over the past 10- 11 quarters. You exited July with 156 days of inventory on your balance sheet. That is down sequentially. Channel inventories continue to be below your target of six to seven weeks. How do you anticipate the trend on your inventory this quarter? Any signs that customers are starting to replenish inventories, or are they still conservatively ordering to their consumption trends?
From an overall strategy perspective, we have talked about this going on almost two years. We have been consciously adding inventory back on our balance sheet.
And actually have started to ship into the channel to add inventory. But the competing dynamic is the revenue growth rates and the sell-through activity is accelerating at a faster rate than folks anticipated. So despite adding to our, I think we ended Q3 with record gross inventory levels. We will expect to add more inventory in Q4, but I still expect days to even likely stay flat. On the channel, same thing. We talked about in the quarter call that we added inventory into the channel, but the sell-through is accelerating so fast, we are also still keeping below our sort of six to seven-week target range. Again, we will, as we have planned, add more inventory into the channel-
but I expect we will still end up in and around that slightly below six weeks. That a little bit is an indicator of how much end quarter business is being generated given the steep demand we are seeing across the end markets.
Yes.
We will continue to try to optimize, just given where we are in the cycle and demand, but I do not expect that we will see an increase in days or weeks despite our efforts. That said, I think we are pretty well-positioned from a supply perspective. Sorry.
Typically, as we move from the bottom of the cycle through the positive parts of the cycle, the first phase of goodness is customers depleting excess inventories, and then they start ordering your products in line with their consumption trends, which is kind of where I think you guys are now. The next phase of growth is when your customers feel comfortable about their demand signals and keeping higher levels of inventories on their balance sheet. We call that the replenishment cycle, right?
Right.
You are right now at a pace where you are shipping into your customers' consumption trends, but are you already starting to see that second phase of growth where customers are starting to replenish inventories, or is that still in front of the ADI team?
I think it is actually a bifurcated story, because I do not think we are actually shipping into consumption across our full customer portfolio. If you look at the three super cycle areas I mentioned before, ATE, A&D, and data center?
Yes.
If you look probably where we are today, I would say we've already gotten to what would've been that historical consumption line and are probably above it.
Okay.
But for the rest of our businesses,
Yes
we are still either at or double digits below. In industrial, for instance, greater than 50% of our industrial sub-segments are still double digits below that consumption line. That's why I say, and when I look at that and I get asked this question all the time, where are we in the cycle? When I look at it, one of the things that's important is for the three super cycles where we're clearly above the consumption line,
That's right.
that is visible end demand.
Yes.
We see data centers being built. We can see what's going on in the Aerospace & Defense world. ATE, although indirect, is directly tied to the
That's right.
amount of testing required for the new, hyperscaler build-outs. Those three, despite being above consumption, I do not believe are in an inventory building period.
Right.
They are trying to get as much as they can to build what they're building today. I don't think we've even finished getting back to consumption broadly, because so much of our business is below that level. Then I would agree, at some point, you'll get the OEMs and end customers, their memories will jump back and say, "Hey, we couldn't ship product because we didn't keep enough inventory in our books, and we will want to put some safety stuff." We have not really seen any of that behavior. I would say that's not been a primary or even a secondary driver of our growth yet. You see it. We can't keep inventory in our channel, and we're not seeing it on our customers. We do. My team looks at our 250 customers, and we track their inventory levels-
Yes
and their revenue levels against ours, and the signals are usually pretty clear when they're starting to put inventory on their balance sheet or deplete it. We only got to depleting inventory in sort of where we feel like we don't have any other inventory challenges, in the last sort of three quarters.
That's right.
Right?
That's right.
One of those cycle signals was we saw the broad market come back, as I mentioned. I also think part of the dynamic was the broad market may have been the last ones to get some of their late allocations from the prior shortages, so it took them a bit longer to work it off. A year ago, we were talking about automation struggling. We think that inventory has been worked off. In the last year and a half, we have been seeing a work off of inventory in our VSM piece of our auto business.
Yeah.
We actually think that that business will come back to growth in the current year. So both from an idiosyncratic ADI-
we think we have got a tailwind, and also from where we are in the cycle. Which, again, is why you heard Vince talk about on our earnings call, a level of confidence that we can continue to grow, in double digits into 2027.
Yeah. Let us take a step back and let us actually look mid to longer term. You talked about the potential, Vince talked about the potential for double digits growth profile in fiscal 2027. But if I look back at the Analog Devices team over the past 20 years, you have grown your revenues at a 10% - 11% CAGR, about 30% faster than the semiconductor industry. Earnings and free cash flow per share growth at a 12% - 15% CAGR over that period of time. Yeah, some of the growth has been inorganic, but majority has been organic, right? I know the team's prior revenue growth targets was like this 7% - 10% revenue CAGR, $15 of per share of earnings power in fiscal 2027, which by the way, you guys are going to blow past that if I look at ours in consensus estimates.
Given the leverage to the data center, the idiosyncratic data center, aerospace and defense, we will talk about energy infrastructure a little bit, continued content gains in automotive, physical AI, how does the team think about the forward revenue growth and earnings power profile of the company? Not just fiscal 2027, but how should we think about that over the next few years?
Certainly, we have had pretty significant growth well above our model for the last several years. We expect that in this upcoming year, we will have double-digit growth again. If you go back and look at our business back to, say, 2018/2019, we have essentially run at a CAGR about that 10.5%.
Yeah.
That has outperformed. Do we think with what we see from bookings backlog design wins that we can continue to outgrow the market? I do. The other thing that is important in there, and you talked about the leverage, is we have been super disciplined from an execution perspective. We talked about coming out of the downturn that we would be very judicious in how we deployed capital. We have made those decisions in a smart way. Obviously, first call for us has been R&D. It will continue to be R&D, and which are the most important high-leverage projects, and we have been very disciplined adding expense elsewhere. So you have seen that in the EPS and cash flow numbers coming through.
Yes.
I think the resiliency of our model will continue. So I think we are very well-positioned. Again, I worry a little bit just because the macro continues to be uncertain.
We are under a ton of cost pressure and inflation pressure. I do think one of the things that has helped us has been disciplined execution. Obviously, we were maintaining price better than most of our peers.
Yes
In fact, have had two price increases. The third thing that's really important, I'd love to highlight this, because we did a very large acquisition a few years ago and promised the world that we would achieve $1 billion of synergies by 2027. I'm happy to say I expect we will exceed the $1 billion, and I think we will close out 2026 with near $700 million worth of synergies. Those things give us confidence that we can continue to be a double-digit grower.
When we look at our companies and we think about the growth profile, the potential growth profile, the one thing that we always look at is your design win pipeline. I think it's a strong forward indicator of the revenue growth potential. The Analog Devices team, you guys grew your design win pipeline by double-digit percentage in fiscal 2023, fiscal 2024, across all end markets. 20% growth in your design win pipeline last year, fiscal 2025. How is fiscal 2026 tracking so far, and what areas of your portfolio or end market exposure are you seeing the strongest expansion and design win momentum?
I agree with your sentiment. Best way to think about where we're headed the next three to five years is what design win looks like. We're on track in 2026 to exceed the growth rates we saw in design wins in 2025. I would say, as you'd expect, we're seeing outsized growth numbers in data center, in aerospace and defense. I would also say broadly across our portfolio, design wins will be up over the levels we saw in 2025.
Got it.
We feel really good about what we're doing from an engineering perspective and how we're deploying our capital, given what we're seeing from the design win perspective.
Let's talk about some of your end markets. I'm going to start off with your largest end market, which is industrial. It's about 50% of your total revenues. I love industrial because it's this diverse set of different businesses, factory automation, healthcare, automated tests, instrumentation and tests, energy infrastructure, Aerospace & Defense. As you've pointed out several times, one of the strong dynamics among several, which separates ADI from its peers, is your strong exposure to Aerospace & Defense. Strong mixed-signal, RF, millimeter wave, DSP, compound semiconductor products. This was a growth area for ADI during the last downturn, continues to drive a strong growth profile. Last year, this business, when we were at this conference, I think was just about breaking through $1 billion annualized revenue run rate.
Where is A&D currently trending on an annualized run rate basis, and what types of programs is the team leveraged to? Is it SATCOM? Is it radar, missile systems, et cetera? What's the growth outlook look like over the next few years?
As you said, a year ago, we were talking about that business just getting to $1 billion. If we look at the Q3 exit rate for that business, it is nearing $2 billion. Obviously substantial amount of growth. The important thing to think about is a couple things. One is our Aerospace & Defense business, that includes what we are doing both from a government contracts as well as our commercial off-
That is right.
the shelf products, where we are engaged in satellites, both, whether it is the LEO satellites or the geosynchronous satellites. It is in drones. It is in missile defense systems. It is a pretty broad portfolio, and it obviously leverages a lot of the technologies that we have built and technologies we have acquired that you mentioned. I do not need to relist them. We feel very strongly that that business can continue to grow double digits for the next decade. When we look at the external signals, the U.S. is asking for a 50% increase in their defense spending.
Right.
The Europeans are going to all spend some percentage of their GDP. I think the opportunity is real, and we are very well-positioned across our portfolio in Aerospace & Defense.
Within industrial, the other idiosyncratic driver has been automated test equipment, what you call ATE. Another strong segment, as you mentioned. You are benefiting from the demand of these automated test equipment systems that test NVIDIA GPUs, Broadcom XPUs, Intel, AMD, CPUs. Your components go into automated test equipment that tests these new high-speed memory architectures, like high-bandwidth memory, DRAM, that supports, let's say, an NVIDIA-based or AMD-based GPUs. Here, the team leads with its leadership position in high-speed data converters and high-speed precision amplifiers. It is trending up strongly year-over-year. What is the annualized revenue run rate of this franchise right now?
Exiting Q3, our ATE business' annualized run rate was just about $1 billion. As we have talked about in the past, we have an incredibly strong share position in ATE. One of the outcomes of this increase in high-bandwidth memory and high-performance compute is the complexity of the testers increases. So each generation of more complex testers tends to drive more ADI content. So, a little bit similar to the auto story, where in addition to getting share, we are also getting increased content as the complexity of the testers increase. I think that that is another area, given what we see in the forecasts from an AI/data center-
That is right.
infrastructure build, that is another business we think that has tailwinds, and gives us an opportunity for that above-market growth.
Sticking with industrial, the diversity of the business that I talked about before, an emerging end market segment within industrial is your energy infrastructure and energy systems business. This has tie-backs to data center and all of this AI build-out. The team has, I think, sized the energy infrastructure/systems market is on a $500 million annualized run rate for this year. Energy storage systems management and monitoring is extremely important for data center infrastructure. You leverage here your expertise in battery management solutions. What other building blocks in the portfolio do you leverage to target this fast and emerging segment within industrial?
Yeah. So obviously it is a really important sector, the energy management, because back to the infrastructure, everything seems to circle back to the data-
That's right.
center infrastructure. The power constraints are going to continue to be one of the challenges for that world. So the ability to manage energy, including energy storage systems, is going to be critical to the ability to get renewables onto the grid to manage that distribution. So our ability to work with the energy businesses is critical. BMS, obviously, our reputation is we have the best BMS systems in the world. So not just in cars, but if you're building batteries for that structure. If you look at our broader signal chain, one of the things that the energy companies want and need from us is our ability to sense how power is being used, what's happening in the infrastructure, and give them telemetry and data so that they can enhance the efficiency of their operations.
That is the broad part of our sensing and signal chain, where we are leveraging our core franchise products in conjunction with our intelligence and algorithms to be able to help the energy providers manage and deliver power efficiently. Because there are finite places to build and get more energy, so the holy grail is getting more efficient in how we use the power that is available. Which is why we feel really good about where we are positioned there. It also gives us a lot of insight in the power from, essentially, you heard Vince talk about this on the call, going from power coming out of the grids all the way to the power going into the GPU/XPU. That is really helpful for us when we are working with the large data center customers.
Yeah, and that is a good segue into my next focus area, which would be your core, what I call your core data center and accelerating compute segment. You have articulated your data center exposure, power delivery, power management products, optical networking control products. At around 80% of your total comms infrastructure business, this data center portion of it is up 100% year-over-year. It is about a $2 billion annualized run rate, split 50% power, 50% optical. On power, obviously, strong position in power delivery. Help us understand the breadth of the team's portfolio here. You and Vince have talked about power delivery from grid to chip. Help us understand the diversity of products and solutions that you attack this particular segment of the market with.
On the power side, if you think about my comment earlier on the energy management, this is getting power from the power distribution companies into the data center. We obviously are helping get the power to the wall of the data center.
That is right.
And then as you go into the data center, the three ways I think about it is there's the power management piece. Think about our hot swap capabilities and our products that do that. The other piece of it is power conversion. Because you've got high voltages coming into the data center that you've got to step down through intermediate, and then you've got to get to that stage two, which is sort of delivering the power of the chip. So I think we have historically been strong in the first two pieces of that, and an area we're very focused on is that last micron, which is the power delivery into the chip.
That's right.
And that's actually where the Empower acquisition we did.
really important, because that gives us an opportunity to have a product in the IVR that is a vertical power solution, which improves the efficiency and effectiveness and helps solve some of the complicated challenges with power delivery that is lateral. Because that chain from the lateral power into the chip, you have a lot of power loss. You're a little bit better when you build underneath. But that is also very complex because of the dimensions and the science. You've got thermodynamics and noise and diffusion issues you have to address that the IVR addresses. So we think that's a really important piece for us to be able to play, and it helps close our loop all the way from grid to the chip.
From a hyperscaler perspective, AI, however you want to describe this AI build, if you can, our analysis would say we can save 15%+ power loss. If you start looking at the size and scale as we move to gigawatts worth of additional data centers, that's a lot of power savings.
Yes.
We think that's a really big opportunity for us. Actually, I think that when you look at the long-term opportunity in the data center, probably for us, the biggest growth opportunity is in that stage two power delivery. The second piece for us in the data center is really what we're doing in the optical space.
Here, our primary position is in the optical control system, controlling the lasers. We have held a leading position there for years, and we've tended to set the standard. Like the power, which we're seeing increases in voltages as we think about the transition to 800 volt and what that means, we're very well positioned. What's also happening is in the data center, the speed with which data has to move across the data centers and across servers continues to accelerate. We've had very strong positions in 400 volt and 800.
Yes.
We're already shipping a 1.6 Tb solution into the market.
We're also, and I think we may have even talked about this last year, we've been working and our customers are already working on what is the USB 3.2 speed.
Yeah.
Each generation of the increase in speed also gives us opportunities to gain share and to get some more content. That control, and then the other piece that is a growing area, and I think Vince talked about this on the call, is in the optical control systems. I think there's a number of those areas where our core products continue to advance to be able to handle the faster speeds, just like we're handling the higher power consumption. We feel like we're very well positioned there. Again, at a $2 billion run rate business, this is really significant for our growth going forward.
The team has talked about, on the last earnings call, $1 billion-$1.5 billion in analog SAM opportunity per gigawatt of data center build-out. How do we think about this split between optical and power?
I'm not sure. I don't know if I have an exact split, but I would say right now, those two businesses for us continue to grow at pretty similar rates, right? We've been talking about continued growth.
Yes
100%. I do think, just given the nature of the power growth for us, the growth rates in power for us will exceed the growth rates in optical, because we're going to have an opportunity to get a power position in vertical power-
That's right.
that is significantly better than the power position we have today in lateral power. So, over time, I think that shift moves away from the 50/50, but I do think given our strength in both positions, they're both very solid long-term growth opportunities.
We talked about the energy systems, and I don't think you guys actually lumped this into your definition of data center, but there's so much focus on energy systems, the capabilities that are needed that you deliver from your battery management and monitoring systems. So I would assume this doesn't add to the data center, but it sort of biases the Analog Devices team more, again, towards that sort of power part of the market opportunity relative to maybe, let's say, optical.
Yeah. It does, as you said, the energy management rolls up currently into our externally reported industrial.
Right. That is right.
However, it is the knowledge that we have gained-
in having those relationships that helps us fully understand the power tree coming from grid into the chip, which obviously helps our data center team. Look, there are a lot of challenges at higher voltages, whether it is thermal and/or otherwise, and I think that is a really strong opportunity for us, which is why I said I think the longer-term growth for us is bigger in the stage two power.
One of the things in my mind that has always sort of differentiated the Analog Devices team is you've always approached things from a systems-level approach. One of the things I feel like the market doesn't appreciate is you actually do have. We all tend to think about Analog Devices as high-performance, mixed-signal analog, and power, but you have a very strong franchise in digital signal processing. You have a very strong franchise in software and firmware development, and systems development, right? How is the team leveraging your software, your digital signal processing or DSP capabilities, your systems-level integration capabilities to gain even further traction in the data center end market?
Great question. I would say what we're trying to do, and I think we're doing a pretty good job, is drive a very similar strategy to what we drive in the broader part of our business, which is leverage all of those components.
to be able to deliver system-level solutions. Because what we find, and we're in conversations directly with the large players here, is they come with a really complex problem. They don't want to solve these problems across a bunch of suppliers. They want help from one supplier to bring them an integrated solution that solves their problem. So our ability to leverage what we've built from a core analog, but also from our digital signal processing, and now adding into it software and our own algorithms, gives us an ability to provide a broader end-to-end solution into the data center. So I think that strategy is what will help us sell more solutions, and that has proven to be a really successful strategy across the rest of our business, and I think our teams are leaning in real hard there.
Maybe as a part of that, two weeks ago, the team announced its acquisition of Alif Semiconductor. That's a company that makes specialized low-power microcontroller, compute-focused chip solutions for AI edge use cases across industrial consumer end markets. What does the acquisition enable the Analog Devices team to do that it couldn't have achieved organically or through partnership? Can you just give us a sense of Alif's annualized revenue run rate and gross profitability profile?
Sure. For us, Alif obviously brings us an AI-native platform of heterogeneous compute. What that does for us is it gives us a product immediately on a platform to bring out that integrated solutions that we want to have for our customers. One of the things that is really important is having that core product allows us to bring a lot more of our broader Analog Devices portfolio into our solutions, gives us a significant upside. I think this was a technology that probably over time we could have developed, but where they were, this was a really important time accelerator for us.
Yes. Mm-hmm.
What we have learned is being in early in these new product situations gives us a significant advantage from a learning cycle perspective. Thinking about what does it mean, right? We talked a lot how we were going to benefit from the infrastructure build as a company.
Yeah.
I think if you go back, geez, right when I started, maybe one of my first earnings calls, Vince talked about the bigger opportunity for ADI was going to be AI at the edge.
That's right.
Small language models.
Decision making out at the physical edge. This helps enable that. I try to use examples I can understand. You think about robotics is a great example, right? We already have the best capabilities in the world at tracking, measuring, and sensing physical data. You are able to take that data, bring it into the-
through the signal chain to a compute, which then can make a decision and drive a local action. That ability to sense, compute, and act out at the edge in this physical intelligence world is a significant next generation of AI because it covers a bunch of things. One, if you think about what are some of the biggest challenges, power. Well, having to compute with small language models out where the data is collected-
Yes
lower power consumption, lower latency, which obviously if you are dealing with certain kinds of robots, latency is super important. You have the signals all happening inside small spaces, not going back and forth to the cloud. Third, security, because you are also not transporting the valuable data you are collecting away. We think that this gives us a significant step forward in speed to be able to start delivering those kinds of products. Our teams have been working with the Alif technology and some of our algorithms to see that they are going to be able to do that, and we feel very good.
I think this is an accelerator for us and a place we knew we needed to get to as part of our long-term strategy, and we feel really good about the products they have and, by the way, where they are with customers on their products.
From our financial implications, obviously we have not closed the deal yet, but we do not expect this to be near-term material from either a revenue or a cost perspective. As we start to look out, this probably is a part of our business that ramps more in the sort of 2028 timeframe.
I see. We know the management team from Alif, Syed Ali. He was the founder of Cavium, a leader in compute. Look at their solutions. I was perusing the Alif portfolio of products, and it is microcontrollers, it is microprocessors, it is standards Arm-based, but they have this unique sort of AI acceleration capability built into every one of their MCUs and MPUs. To your point, it sounds like they have a very strong design win profile across industrial, IoT, consumer, wearables type applications.
The other piece of their business that is important also as we look out at the edge is their ability to do sensor fusion.
Yes. That is right. Let us talk about pricing. Pricing has been an important dynamic in the analog sector this year, right? From what we have gathered, I think as you mentioned, the Analog Devices team has rolled out two pricing adjustment actions this year, one in February, another this month. Can you just help us better understand the actions taken year to date, whether they were applied broadly across the portfolio or concentrated in specific product families or in different end markets?
Sure. Just to level-set from a pricing philosophy perspective, Vince and I have talked about this before, our goal with the pricing, and it includes the last two we did, is really to offset the sustained inflation we have been experiencing, right? We have worked very hard over a number of years to try to absorb as much of that inflation through efficiency to maintain our margins, but it has just persisted too long. The first price increase that we did was largely targeted at our channel.
Again, our price increases are cost recovery. There is no margin stacking. We are not doing this to increase margins. We are just doing it to try to keep our margins stable at the existing levels. This most recent increase, we have continued to see, early on, the inflation was in some obvious places, whether it was transportation costs or gold costs. That has continued to broaden. We are seeing inflation in wafer costs, we are seeing inflation in assembly pricing, pretty much across the supply chain, which drove the second price increase. That one was broadly across all customers. That, as you described, was rolled out in mid-September. We will continue to monitor the market, the dynamics. Look, the inflationary pressure is not receding. I know, at least in the States, there is a rate increase that hopefully will drive some moderation, but we will continue to watch that.
We will also continue to try to be as efficient as we can to absorb some of that cost. Again, our philosophy going forward will be to recover costs when we need to to maintain margin.
Yeah, because to your point, and we cover guys like GlobalFoundries. Our Asia team covers companies like TSMC, UMC, and SMIC, and it does feel like things, supply-wise, capability-wise, aren't getting better as we step into 2027. It looks like some of your foundry partners are getting set to potentially increase wafer pricing further. Based on what you said, if input costs, one of which includes wafer pricing, continues to increase, there's potentially maybe a likelihood that ADI will have to take some more pricing actions as a result of that, potentially.
Yeah. We do not have any incremental pricing actions planned.
Right.
As I mentioned, our approach will be consistent. We'll monitor and track, and if it looks like it's going to materially have an impact, we would consider an incremental price increase. Obviously, that's not our preferred path. We'd like to see that inflation come down.
As you continue to move through this upcycle, could you just discuss some of the key gross margin numbers that will allow the team to maintain your near-record 70% gross margins into next year?
Sure. Actually, one of the interesting things is, for Q4, we have guided to a record margin.
Yeah.
What is interesting about that is the last time we had a record margin, we were a 53% industrial business.
We are guiding to this record margin at a 50%.
Right.
One of the areas where there is an opportunity to continue to maintain in the face of some of these pressures, or maybe even increase, is mix change, right? If we see a further shift to industrial, which tends to be our overall higher margin business, we could see some upside. Obviously, the revenue curve will be a big indicator. If we see this continued acceleration on the revenue curve, then we will just see better fixed cost absorption, which could give us some benefit. Then the third piece, as you mentioned, would be any potential future price increase. But obviously, our preference would be to have costs moderate. But those are the things that will help us stay at that level. The mix within mix has some impact, but broadly, it is that shift to broader industrial that gives us the uplift.
I think I talked about this in the last call. I'm not expecting utilizations to be a driver
That's right.
of further upside because we're running the factories pretty hot. In fact, we're adding new capacity every day into the factories, adding equipment to available clean rooms, adding test capabilities, and actually getting more capabilities externally, just given the continued growth ramp that we're experiencing.
The team had previously targeted best-in-class free cash flow margins 34%-40%. You just drove 36% free cash flow margins, trailing 12 months, with greater than 100% free cash flow return via dividends and repurchase. Is that still the right sort of range to think about free cash flow generation on a go-forward basis as a percent of revenues in that sort of 34%-40% range on a go-forward basis?
Short version, yeah, 34% - 40% still is the right model. If you look across our peer set, we tend to be in the top part of that.
That's right.
One of the things I think that's important about our free cash flow model and the result is it's really a reinforcer of the operating model we're running. We're continuing to generate significant operating margins, and we're translating that into cash, highly effectively. Our ability to use that to return cash to shareholders through dividends, which have increased for 20-something plus years, and stock buybacks has been critical for us as we look at the overall. There'll be some variations. If you saw at the bottom, we were still very strong free cash flow. We had a record free cash flow 25 this year. We've talked about, we've had some incremental payments for some capacity. We've got some incremental tax payments things.
Yes.
Still feel very good about where we are in that 34% - 40% range. Our goal is, I think, we need to continue driving to the high end of that.
We're just about out of time, but I wanted to see if there are any questions from the audience. If you have any questions, feel free to
Jeff, you don't get to ask questions.
Raise your hand. I did want to ask a question about your mass markets business because you referred to it broad markets, mass market. I tend to view it as sort of your catalog set of products, right? These are products where they can be used in multiple different applications, multiple different end markets, very high performance, and it gets sold to this very long tail of small to medium-sized customers. To your point, these category of customers oftentimes end up being sort of the leading indicator from a sort of cyclical perspective because they typically tend to move first as it relates to pulling back or starting to add more purchasing capabilities. Can you just help us understand roughly the sizing of the mass market, broad market business, and the gross profitability profile? Because my sense is gross profitability for mass market is actually quite rich.
I'm going to give an answer, and then Jeff will correct me if I'm wrong. I think the broad market's probably about 15% of industrial, and I would say on average, it's above the corporate gross margin average. I would tell you just one point before we stop on this. Our broad market, many of our large customers also buy through the broad market.
I see.
Because we don't dictate where people make
Right
their purchases from us, so sometimes fulfilling it through the distribution channel for them is a preferred choice, and we're fine with that. So some of our big customers that we have direct agreements also buy through what would show up in the channel business.
I see. Back to our discussion on the potential gross margin and the potential for maybe gross margins to continue to inch its way higher. Broad markets falls under industrial, which is one of those categories, non-AI focused, that is probably still sitting below where you were back in the peak of, I don't know, 2021 or 2022, right? As that broad markets business comes back, higher gross profitability profile.
Yeah.
That is a potential for maybe driving slightly higher.
It's interesting because I think we talked about if we get through Q4 where we expect, I think that will be four quarters of where we've seen broad market growth.
Yes.
Still much shorter than what would be a typical cycle.
Right
Before we'd see a turn, which is another one of those data points we look at to decide where we are relative to the cycle.
That's great. Strong setup for 2026. Strong setup for 2027. Look forward to continuing to monitor the execution of the team. Rich, thank you very much for participating.
Thanks for having me, Harlan. Appreciate it very much.
Yeah. Thank you.