Analog Devices, Inc. 2026 Q3 Earnings Call

Thank you, Danny, and good morning, everybody. Thank you for joining our third quarter fiscal 2026 conference call. Joining me today is ADI CEO and Chair, Vincent Roche, and ADI CFO, Richard Puccio. For anyone who missed the release, you can find it at investor.analog.com, along with related financial schedules. The information we are about to discuss includes forward-looking statements, which are subject to certain risks and uncertainties as further described in our earnings release, periodic reports, and other materials filed with the SEC. Actual results could differ materially from the forward-looking information as these statements reflect our expectations only as of the date of this call. We undertake no obligation to update these statements except as required by law.

References to gross margin, operating and non-operating expenses, operating margin, tax rate, earnings per share, and free cash flow in our comments today will be on a non-GAAP basis, which excludes special items. When comparing our results to our historical performance, special items are also excluded from prior periods. Reconciliations of these non-GAAP measures to their most directly comparable GAAP measures and additional information about our non-GAAP measures are included in today's earnings release. References to earnings per share are on a fully diluted basis. With that, I will turn the call over to ADI CEO and Chair, Vincent Roche.

Thank you, Jeff, and a very good morning to you all. As you have seen, third quarter revenue, margin, and earnings all exceeded our outlook, with growth across all of our end markets, led by data center and industrial, propelling us to the first $4 billion quarter in ADI's history. Demand for our solutions continues to grow, supported by robust AI and defense spending, cyclical momentum, and underlying secular content growth across our diversified end markets. Through targeted R&D, we continue to extend the limits of technology performance and accelerate the pace with which we are delivering more comprehensive solutions to our customers' toughest problems. In tandem, investments in our hybrid manufacturing network have enabled us to increase the agility and responsiveness of our supply chain and consistently capture above seasonal growth for more than 2 years.

Now for the rest of my remarks today, I'll focus on how we're helping customers meet unprecedented and still accelerating demand for AI infrastructure and energy systems. The fact that data center capacity is now measured in gigawatts rather than FLOPS and TOPS underscores one of the most defining challenges of the AI era. Power availability has become the primary constraint to further AI progress. Solving this challenge requires more than simply adding more energy, however. It demands a grid-to-chip system-level approach that encompasses both improving the availability and delivery of energy and extracting the maximum computing power from every watt delivered. Now, let me walk you through some of the key elements of our grid-to-chip strategy, starting at the grid where the AI bottleneck begins. As electricity networks become more complex, visibility, efficiency, and resilience are becoming critical challenges.

Customers are turning to ADI's grid monitoring solutions to illuminate the flow of energy across the network, providing real-time insight into voltage, current, power quality, and system health. Our higher-value solutions are helping utilities, energy operators, and infrastructure providers to improve efficiency, reliability, and utilization. An increasingly essential part of the grid and one of the fastest-growing sectors is energy storage. Here, customers choose ADI's industry-leading battery management technology to help maximize usable energy, improve system efficiency, extend battery life, enhance safety, and of course, improve ROI. Expanding and modernizing the traditional grid alone, however, is not enough to keep pace with the speed of AI infrastructure deployment. To accelerate time to power, hyperscalers are increasingly exploring dedicated microgrids, which are opening up additional avenues of growth for ADI.

We believe this trend of localizing power will augment our $500 million-plus energy business, which began inflecting in 2025 and has been delivering accelerating growth this year also. Importantly, our strong and growing positions across both energy and data center make us a more critical AI ecosystem player, spanning the entire electricity value chain from generation, transmission, and storage, to distribution through rack power and ultimately processor power delivery, essentially the vascular system of the data center. Now, once the grid makes contact with the data center, AI's extreme energy and information density requirements make ADI's deep expertise and innovation in high-performance power management, sensing and telemetry, as well as optical connectivity even more critical.

Let me start with our optical franchise as I begin to unpack for you how we're growing our data center business and opportunity by helping our customers resolve the tremendous challenges of energy and information density. When we think about the journey of data through the infrastructure, there are two critical pathways, the data path of electro optics and the control path, which guides, optimizes, and ensures the integrity of the data path. Our focus is on the control path, where we've been setting and extending the industry's performance envelope for decades. Today, the complexity of efficiently moving data at ever higher speeds within and between racks and across data center campuses is growing exponentially.

Customers are increasingly relying on ADI to provide essential timing, power management, data conversion, monitoring, and control capabilities that enable lasers and transceivers to operate with precision reliably, efficiently, and at the necessary scale for AI workloads. As customers seek to further increase the amount of optical lanes, signal bandwidth, or both, to accelerate network speeds from 800G to 3.2 terabits per second, we believe that we're very well positioned to benefit threefold from unit growth in pluggables and coherent light modules, increasing BOM content, and greater share as these transitions unfold. As new architectures such as optical circuit switching and co-packaged optics gain traction in next generation large scale AI clusters, complexity expands even further, and our long-term opportunity continues to grow.

Based on current design wins and customer commitments, our OCS revenue is poised to approximately double this year, and we're targeting a similar level of growth in 2027. In the nascent CPO space, which we view as a SAM expander, the criticality of ADI's precision control technology further increases as thermal and serviceability challenges rise. In short, the combination of market growth, expanding content, increasing share, and differentiated value creation across data center optics reinforces our confidence that this segment will remain a strong growth vector for ADI over the coming years. Now let me turn to our power franchise. The need for customers to convert and deliver precise increasing levels of power at the rack and compute layers efficiently and safely is driving continued broad-based growth across our portfolio.

Customers are leveraging ADI's products and solutions to push for greater than 98% conversion efficiencies, multi-kilowatt power delivery with peak power levels up to two times the rated load, and comprehensive protection, telemetry, and fault recording capabilities that enhance system reliability and maximize uptime. To put just one of those differentiators in context, the 1% difference between 97% and 98% efficiency may not sound like very much, but a 97% conversion efficiency loses roughly 50% more energy through heat than a 98% solution. Over time, of course, that difference compounds in terms of the need for additional cooling infrastructure, stress on equipment, and operating costs. We believe our opportunity will continue to grow substantially as power density demands of AI clusters continue to increase. The industry's architectural transition toward 800 volt DC power distribution, for example, plays directly into ADI's power management expertise and portfolio.

We're seeing a significant design and uptick for our protection and 800 volt to intermediate power conversion technologies, which can deliver 20 kilowatts of power at industry-leading power densities exceeding 2.5 kilowatts per cubic inch. At the intermediate to core conversion layer, which is one of the fastest and largest growing analog opportunities in the AI era, our combination of advanced power conversion, intelligent system control, and real-time telemetry is critical to achieving the necessary power density, efficiency, and reliability requirements for next generation processors to operate at 6,000 amps and sub one volt. Our Empower acquisition further enhances ADI's vertical power story by enabling us to take power into the processor package itself.

In large scale AI deployments, these architectural advantages can reduce compute power consumption and temperature by approximately 10%-15%, which equates to roughly $30 million in annual savings in a one gigawatt data center. As with optical, our power pipeline is growing rapidly, and the direction and rate of our R&D investments reflects our belief in the size of the SAM opportunity before us and our confidence that data center power can remain a strong growth vector for ADI over the coming years. In summary, we believe the architectural shifts underpinning the evolving AI era are increasing ADI's role as a critical partner across the grid to chip ecosystem and driving extraordinary opportunity. Our current assessment is that our 2030 data center and energy SAM has more than doubled from what we had envisioned just one year ago.

This dramatic expansion is not simply a function of increased AI infrastructure CapEx. It reflects the impact of new markets and architectures that require orders of magnitude more analog content delivered via higher value solutions. Now stepping back to frame this growth on the larger landscape of ADI's continued evolution, grid to chip is but one facet of the first generation of AI, characterized by applications largely focused on data centers. The ATE growth we've recently spoke to on these calls is yet another facet. As great as the impact of generation one AI has been so far for ADI, however, we continue to believe that the bigger prize may be in the second generation as AI extends its reach from the data center to the physical world in the form of pervasive robotics, digital health, autonomous mobility, and so on and so forth.