Earnings Call Transcript - ARM Q1 2026

Jeff Kvaal, Vice President, Investor Relations

Thank you very much, Sharon, and welcome, everyone. This is our first earnings call of fiscal year 2026. On the call today are Rene Haas, Arm's Chief Executive Officer; and Jason Child, Arm's Chief Financial Officer. During the call, Arm will discuss forecasts, targets, and other forward-looking information regarding the company and its financial results. While these statements represent our best current judgment about future results and performance, our actual results are subject to many risks and uncertainties that could cause results to differ materially. In addition to any risks that we highlight during this call, important risk factors that may affect our future results and performance are described in our registration statement on Form 20-F filed with the SEC. Arm assumes no obligation to update any forward-looking statements. We will refer to non-GAAP financial measures during the discussion. Reconciliations of certain of these non-GAAP financial measures to their most directly comparable GAAP financial measures can be found in our shareholder letter, as can a discussion of certain projected non-GAAP financial measures that we are not able to reconcile without unreasonable effort and supplemental financial information. The shareholder letter and other earnings-related materials are available on our website at investors.arm.com. And with that, I'll turn the call over to Rene.

Rene Anthony Andrada Haas, CEO

Thank you, Jeff, and welcome, everyone. We began fiscal year 2026 with strong momentum, fueled by the insatiable compute demands of AI. From smart centers in homes and factories to the world's most advanced AI supercomputers, AI workloads are being deployed everywhere. This is driving unprecedented demand for compute that's not only performant, but also energy efficient. Arm is the only compute platform built to deliver AI performance across the full spectrum of power and performance from milliwatts to megawatts. As a result, Q1 was our highest revenue quarter — second highest revenue quarter at $1.05 billion. Royalty revenue reached $585 million, up 25% year-on-year with strong momentum across all of our end markets. Licensing revenue was $468 million as companies continue to make Arm the AI platform of choice. Custom silicon on Arm is driving unmatched AI scale in the cloud. More than 70,000 enterprises now run AI workloads on Arm Neoverse data center chips, a 40% increase year-on-year and a 14x surge since 2021. Arm Neoverse CPUs now power the most important AI infrastructure in the world, including NVIDIA Grace, AWS Graviton, Google Axion, and Microsoft Cobalt, among others. Driven by performance and efficient compute, for example, NVIDIA Grace Blackwell is 25x more energy efficient than the previous x86-based system. We expect the market share of Arm Neo-based chips to top hyperscalers to reach nearly 50% this year. We are also seeing AI moving to the edge, where the need for local real-time intelligence is enabled by Arm's efficiency and scale. Our compute platform delivers a unique combination of AI performance acceleration and energy efficiency, including the Ethos-U85 NPU for enhanced image recognition and our v9 CPUs with scalable matrix extensions for accelerating language models, which can boost the performance of these models while limiting power and latency overhead. Technology leaders, including Apple, Samsung, and MediaTek, are integrating these AI capabilities for faster, more efficient AI on premium smartphones. AI workloads are going local and Arm is making it possible. Our leadership in AI is amplified by our unmatched software developer ecosystem. Over 22 million developers, more than 80% of the global base build on Arm. This ecosystem is a powerful flywheel. More developers mean more software availability, which in turn drives more demand for our compute platform across every market. Our compute subsystems, CSS, are helping customers move faster and the demand has exceeded our expectations. Our first generation of CSS is now in market with 5 customers and is delivering double the royalty of Armv9. We signed 3 additional CSS licenses this quarter with existing CSS customers, including 2 for the data center and 1 for PCs, more than doubling our CSS licenses from a year ago. Recently, Xiaomi launched the XRING O1 and Samsung launched the Galaxy Flip 7 using the Exynos 2500, both based on the latest Arm compute subsystem platform. These subsequent generations of our CSS platforms deliver even greater value, functionality, and time to market and bring the highest royalty rates we have seen to date. This includes the launch of Zena CSS, a platform optimized for AI-driven automotive workloads like autonomous driving. Also in the quarter, a major smartphone OEM has committed to our GPU platform to accelerate graphics and AI in multiple generations of their flagship smartphones through 2030. Our business relationship with SoftBank has expanded to help them build towards their greater, broader AI vision. We are continuing to explore the possibility of moving beyond our current platform into additional compute subsystems, chiplets, and potentially full-end solutions. To ensure these opportunities are executed successfully, we have accelerated the investment into our R&D. These investments include expanding engineering delivery across multiple levels, adding to the already significant product investments we have made to date. Momentum behind the broad CSS adoption and increased demand for AI compute on Arm are driving a powerful growth trajectory for the company. And with that, I'll hand over to Jason.

Jason E. Child, CFO

Thank you, Rene. We have started fiscal year 2026 with another strong quarter. Total revenue of $1.05 billion was our second successive quarter over $1 billion, our best Q1 ever and above the midpoint of our guidance range. Royalty revenue grew 25% year-on-year to a Q1 record of $585 million. Royalty revenue is growing across all target end markets, including smartphones, data center, automotive, and IoT. Smartphones grew at an order of magnitude faster than the market given continued uptake of flagship smartphones based on Armv9 and CSS. Licensing and other revenue decreased 1% year-on-year as expected, following a very strong Q1 of FY '25. Rene mentioned our progress with 3 new CSS deals. What is unique about these deals is that we closed on our first 3 opportunities to upgrade existing CSS customers into next-generation platforms. Our next-generation CSS platforms deliver even greater value, functionality, and time to market and bring the highest royalty rates we have seen to date. This gives us confidence in further investment in the platform. Also contributing to our license revenue, ACV, and RPO growth was a multigenerational GPU deal with a leading smartphone OEM, and SoftBank expanded its IP licensing and design services agreements with us. Licensing revenue varies quarter-to-quarter due to normal fluctuations in timing and size of multiple high-value license agreements and contributions from backlog. As always, we recommend that you look at annualized contract value, or ACV, to best understand the underlying licensing growth rate. ACV in Q1 was up 28% year-on-year. This is well above the high end of our recent run rate of low teens and our long-term expectation of mid- to high single-digit license growth in part given the new licensing deals I mentioned. We have not changed our long-term view of licensing growth of mid- to high single digits. Remaining performance obligations or RPO was flat sequentially as the new licensing deals offset revenue we recognized from licenses signed in prior quarters. As you know, Arm's revenues today come from technology developed years or even decades ago, and our costs today are investments for future revenue streams. In the first quarter, R&D spending led our non-GAAP operating expenses to $619 million. Operating expenses were slightly lower than expected as the timing of some expenses will now fall into Q2 of fiscal '26. The CSS royalty rate increases that we've outlined are one example of the return from previous R&D expense. This translated to $412 million of non-GAAP operating profit and non-GAAP EPS of $0.35, which was above the midpoint of our guidance range, inclusive of a $0.01 FX headwind. Let me spend a moment on the current tariff and macro climate. We continue to expect a limited direct impact on our royalty and licensing revenues given current conditions. We have less visibility into the indirect impact on end demand. The continued uncertainty reduces near-term visibility on royalty revenue. In licensing, customers have historically invested through near-term slowdowns given lengthy chip development timelines. Turning now to guidance. Our guidance reflects our current view of our end markets and our licensing pipeline. For Q2, we expect revenue of between $1.01 billion and $1.11 billion. At the midpoint, this represents revenue growth of about 25% year-on-year. We expect both royalties and licensing to be about flat sequentially. Consistent with Q1, we are accelerating investments in our next generation of technologies. We expect our Q2 non-GAAP operating expense to be approximately $655 million. This includes the impact of the Q1 expenses that will now fall into Q2 plus FX. We expect non-GAAP EPS to be in the range of $0.29 to $0.37. We have high confidence in healthy growth in the coming year and in years to come. Our confidence stems from our visibility into customer design pipelines, contracted royalty rates, rising demand for custom silicon and AI from cloud to the edge. We expect to continue pressing our advantage by investing aggressively in R&D to support our customers and partners, capture our opportunities, and ensure AI runs on Arm. With that, I'll turn the call back to the operator for the Q&A portion of the call.

David O'Connor, Analyst

Rene, you mentioned in your script expanding and moving into full-end solutions. And when we look out there in the market, there are a lot of reports that Arm is potentially entering into things like the ASIC market with partners, which itself is a lot of execution risk for Arm. And also, my question is why do you think you guys would be successful in such a market when others are struggling? So I guess my question is, is there anything that you can share with us today around Arm's strategy in ASICs moving to these full-end solutions?

Rene Anthony Andrada Haas, CEO

Thank you for the question. Yes, you're right. I have nothing to specifically announce today. What can I say, however, is further integration has been the direction of travel for our companies. One of the things that we're seeing with newer customers such as CSPs and OEMs and also traditional customers is a demand for a better starting point as they develop their SoCs. This is largely driven by the complexity of these chips and the time it takes to develop them. This led us to CSS, our compute subsystem, which as I mentioned and Jason mentioned, have been successful beyond our expectations. Many of the chiplets that are being developed are mostly Arm IP, and we already support chiplet development through our Arm Total Design ecosystem. And with that, we're looking now at the viability of moving beyond the current platform to additional subsystems, chiplets, or possibly full solutions. Now inside the company, we have either inside or access to all the expertise and technologies we would need to design, implement, and have chiplet manufacturing. Personally, I appreciate the complexity of this, having lived this in multiple semiconductor companies in my career. And among the leadership team, we also have comprehensive experience in this area. So when we look at what's going on inside the market today, both in terms of the direction of travel of delivering complex chips and Arm being the only compute platform that can provide a solution from the smallest devices to the largest data centers, milliwatts to megawatts, we're in a very unique space to provide solutions in a way that no one else can. And as a result, we're looking deeply at those possibilities.

Vivek Arya, Analyst

On the royalty side, it grew 25%. I think last quarter, you had suggested 25% to 30%. So just which end market drove a little bit of that delta versus your expectation? And then I think you had also suggested that royalties could grow 10% to 15% sequentially in Q3 and Q4 also of this year. Is that still kind of the trend line we should be expecting?

Jason E. Child, CFO

Vivek, this is Jason. I'll take that. So on the royalties in the quarter, yes, we came in very close to our forecast, within 1% or so. However, I think we talked last quarter about a range. I would say we were in the lower end of that range. And I don't know if I'd say there's weakness, but maybe the growth wasn't quite as strong in the smartphone sector as we had expected. As mentioned in the call, we still grew at an order of magnitude faster than the market, but the market was in the low single digits kind of growth rate. I think that was a little bit slower than folks expected. In terms of the forecast for the rest of the year, I don't expect too much difference. Maybe this smartphone impact we'll have to keep an eye on as we go later in the year. But overall, I expect royalties to be pretty close to where we were from the previous quarter.

Joseph Michael Quatrochi, Analyst

You talked about your Neoverse share top hyperscalers expected to reach nearly 50% this year. Any help in the context of what that was last year? And how do we think about just the mix of those workloads running internal versus external?

Rene Anthony Andrada Haas, CEO

Thank you for the question. I think if we look back to a year ago, one of the significant changes that we've seen has been where our share last year was probably sub 20%, call it 18%, and now we're looking to approach nearly 50%. It's a combination of two things. It is what I would call general-purpose workloads where Arm has been taking share versus conventional x86. This is through Graviton and Google Axion and Microsoft Cobalt, those share gains continue. In addition to that, you also have these AI workloads running specifically for training and/or inference, which in previous generations were on the NVIDIA Hopper generation, which is connected to an external x86, now moving to Grace Blackwell, GB200, and then GB300, which is an integrated Arm design along with NVIDIA Blackwell next-generation GPU. So the combination of the growth in AI data centers, also the fact that we are going from essentially 0 share with Hopper to almost exclusive share with Blackwell, and then when you combine that with our growth in the conventional data center market, that's what's putting us close to a net 50%.

Mark John Lipacis, Analyst

I had a question for Jason and maybe a bit of a quick follow-up for Rene. Jason, you mentioned FX was a $0.01 impact to EPS for this quarter. And then you said that the out quarter guidance included the impact of FX, but I don't believe you quantified that. I was hoping that you could just remind us how much FX is expected to impact EPS for the next quarter? What's your hedging strategy and the rule of thumb of how to think about FX going forward? And then, Rene, if your Neoverse chips are going to get to 50% this year, that suggests 50% share still for x86. What is the reason that customers stay with x86? Does your share asymptote somewhere below 100% at some point in time? Or how should we think about ultimately where you guys could get?

Jason E. Child, CFO

Thanks, Mark. I'll answer the FX question quickly. So yes, $0.01 impact on the current quarter. We expect approximately $0.01 per quarter impact over the next 3 quarters for the balance of the year. The way our FX strategy works is we basically start the hedging process at the beginning of the year. Of course, we collect all of our revenue in U.S. dollars. About 2/3 of our OpEx or so is in euros and pounds. We make estimates on our expenses. In these cases where we're maybe starting to increase some of our OpEx related to some of the opportunities that Rene expected, we have some unhedged portion, and that's why there's a little more impact than what we've seen in the past. But overall, for the full year, think something like $0.04 or so.

Rene Anthony Andrada Haas, CEO

Yes. Regarding the question in terms of where could the market share go, a couple of things make us very positive about the share gains continuing beyond 50%. One is the fact that the Arm architecture allows for a high degree of customization and a high degree of unique design capabilities that could be done at the chip level, at the blade level, and even at the rack level to greatly change the total cost of ownership, but maximize overall performance because customization is very beneficial in those cases. So huge benefit there because of our unique position. Secondly, as the share gains continue across the AI data center, advantages exist to keep a single software stack across a single CPU architecture that simplifies management of the overall software of the data center. That's another intangible that could help us. In general, we're very pleased about getting to the number that I just described, and we think it can grow beyond that. We have high confidence.

Andrew Michael Gardiner, Analyst

Jason, I was interested in the points you were making on the significant step-up in ACV in the quarter. You mentioned both the GPU signing as well as additional licensing with SoftBank. Can you give us a sense as to the magnitude between the two of those drivers? And in particular, on the GPU side, in the press release, I think you said a major smartphone OEM in your prepared comments, I feel like I had the major OEM. I just wanted to clarify that. Just any color you can provide behind those two drivers would be great.

Jason E. Child, CFO

Sure. First, I would say it's a major OEM. That's what you should take away. In terms of the other question? The ACV step-up, again, the 3 big CSS deals, which, as I said in our prepared remarks, were exciting because this is actually the first chance for someone who's used the CSS now to decide if they want to adopt the next generation. To sign those deals is a pretty big deal and a verification or testimony to the value that CSS is providing. While that provided a significant amount of incremental license revenue, only a portion of that is booked. Part of it goes in RPO and of course, part of it is booked in the quarter. If you were to look at all those rest of the world deals, that would take ACV up to, call it, 17-ish percent, and then, of course, this large step-up in the relationship with SoftBank and our custom design services relationship and license with them, that's what gets you up to the kind of the 28% level. By all dimensions, we've seen an acceleration from both the rest of the world, our customers across all geographies, as well as specifically with SoftBank.

Joseph Michael Quatrochi, Analyst

Okay. Understood. And just quickly a follow-up on the CSS signings. You mentioned the second generation will generate an even higher royalty rate than the prior one. Can you give us any sense as to the magnitude of that change?

Jason E. Child, CFO

We've said in the past that CSS was roughly — if you think of it — kind of the best example of flagship or premium mobile, it used to be that a v8 was in the 2.5% to 3% of ASP and then a v9 was roughly 5%, so it roughly doubled. CSS doubled that to roughly 10%. The new CSS deals are north of 10%. It's too early to say exactly what the percentage is. But I think in the past, when we launched the CSS program, we thought that CSS probably was — at 10% — getting close to a ceiling. The good news is we're seeing that it's not the ceiling. We'll provide more clarity as more deals get signed and as we start to guide to future periods. But make no mistake, this is definitely higher royalty rates than what we had expected and forecasted in the out years. So it's definitely good news.

Vijay Raghavan Rakesh, Analyst

Rene and Jason, just going back on the same question on the ACV. With SoftBank expanding the license deal there, can you give us some more color on how that looks? I think this is probably with the Stargate for next year, but as you look at that — as you guys do an accelerator there, are you going to benefit from the full ASP? Or will it be a licensing deal? Can you give us some color?

Rene Anthony Andrada Haas, CEO

Yes. Thanks for the question, Vijay. I'll try to answer your question as simply as possible, although there's some details you asked for that I can't really comment on. At a very high level, Stargate, which is a joint investment venture between SoftBank and OpenAI, is looking to scale up to 10 gigawatts over the next number of years in terms of overall investment. That is a lot of compute, and there's a huge potential for lots of different design opportunities. SoftBank has a broad AI vision. We're looking to help them with that. Again, without mentioning specific products and application spaces, you can imagine in a data center of that size running different workloads around inference, training, and such, and today, all of the Stargate opportunities use Arm as the core CPU; we have a unique opportunity to provide solutions there. A lot of that work has started, but we're not able to provide any specifics in terms of products or timelines.

Vijay Raghavan Rakesh, Analyst

And then going back on the server side, when you look at the 50% share that you talked about, you obviously have some good big customers there between Google and Amazon and, obviously, Microsoft. But you also have another CSP that is potentially starting. Do you expect Arm to be in the market with the merchant CPU? Or will it all be kind of licensed platforms that run internally?

Rene Anthony Andrada Haas, CEO

Yes. Right now, I'm not able to comment on your question regarding us providing a specific product. The market share gains that we anticipate over time are somewhat independent of that strategy. We feel super confident both in the direction of travel of investment in software that's running on Arm and the roadmap that we'll be able to provide the product technology in any form to increase market share.

Kinney Chin, Analyst

This is Steven calling on behalf of Krish. Rene, I had a question for you, kind of big picture. Just over the last couple of weeks, a number of hyperscalers have increased their CapEx targets for this year and also talked about further growth into next year. I'm kind of curious about your initial thoughts on the implications for number one, royalty growth for this year and next year as it pertains to existing Arm processor ASIC type design wins and also looking further out for licensing opportunities for further CSS type wins.

Rene Anthony Andrada Haas, CEO

Yes. Thank you for the question. I think I would interpret this continued increase in CapEx for all of these AI data centers as a very strong tailwind for Arm, both on the technology side in terms of the products that we provide and also with the royalty rates associated with it. As Jason mentioned, we're now on our second generation of CSSs. We talked about CSS as a concept a couple of years ago before we went public, and we're now starting to see the benefits of those hitting our royalties. Over the next number of years, as these newer CSSs with even higher royalty rates hit, we should expect an expansion of the royalty growth on that. Right now, on the overall CapEx side, what we are seeing is still unabated demand for that. One of the broader reasons for this is that AI has one of the unique capabilities that it will touch literally every industry that we know. Over time, there isn't really going to be anything that can't be impacted by it. It's still very early days in terms of how enterprises are broadly using it. I think you're seeing that in terms of the hyperscalers being very aggressive on spending. The evolution of the models supports that the more compute, the better the models get, the better the models get, there's more need for compute. I think for us, it's all good because we're at the heart of compute. As these models evolve and start to move to different parts of the ecosystem more broadly, those are also domains where Arm is already the compute platform of choice. I think it all adds up for good news for us in the long run.

Sebastien Cyrus Naji, Analyst

I just wanted to ask if you could provide any commentary on the Arm China business and in particular, whether some of the loosen GPU export controls we've seen here over the last month could drive a more meaningful contribution from Arm China in the data center versus your prior expectations even 3 months ago?

Rene Anthony Andrada Haas, CEO

Yes. One of the things — thank you for the question. One of the things that we've been pretty consistent on communicating is that for Arm, the China market largely tracks the global market in terms of China really does rely very heavily on the Western ecosystem relative to the software that runs on Arm. So whether it's smartphones or autonomous vehicles or the data center, our traction and momentum in China are quite consistent with the rest of the world. To your direct question on have we seen anything potentially on the new export controls changing anything regarding our business? Not really. The specific H20 release doesn't impact us very much. I would say on that question, no. But again, more broadly, with Arm China, we continue to see consistent growth there that's aligned with the kind of growth we see across the rest of the world.

Jason E. Child, CFO

This is Jason here. Just to add on. China in Q1 was about 21% of revenue. So it ticked up a little bit from the 15% that it was last quarter and about 14% of where it was a year ago. The China business is, I would say, strong and continuing to grow nicely.

Stephane Houri, Analyst

I have a question about the adoption of Armv9. If you can give us the number where we stand now. Last quarter, I think it was about 30%, if I remember well. Where are we standing now?

Jason E. Child, CFO

Thank you, Stephane. Yes, we actually said last quarter, we gave the update for end of year, and it was a little over 30% and that we're now just going to update this on an annual basis. However, still, wait until the end of this year is the next time we're going to provide the financial update or the percentage update. But what I can tell you is that royalties did step up from 18% year-on-year growth last quarter to 25% this quarter. So you should assume that B9 and CSS percentage continues to grow nicely.

Rene Anthony Andrada Haas, CEO

Yes. The thing I would add on to that is, and one of the reasons why we're not communicating to the level that we had in the past, it can be misleading because with version 9, for example, on some mobile phone iterations, we're on our fourth generation of v9 implementations delivered. On each generation, the royalty rate has increased. Combining royalty rates increasing from 8 to 9, but more significantly, each generation of v9 is higher than the previous one. Those generations are on a per annum basis or at least the implementations that use them; the royalty rates will continue to grow faster than v9 adoption. We want to ensure that we communicate that this correlation between royalty rate growth and v9 penetration should be separated out.

Lee John Simpson, Analyst

I'll maybe ask a two-part question if I can get away with it. You called out Ethos and Zena CSS in the prepared remarks. So I guess with Ethos, first of all, it has a good capability set at the edge, but does it make sense to stretch the performance here in time and go beyond that 4 tops level? Where could an MPU like that be applicable long-term? Could it be used in a cloud environment? Is it applicable as a performance level there in time? And maybe just on Zena, interesting that you've had the first deal done already in CSS. But obviously, there's a range of opportunities in the car. So I'm trying to understand, do we think of this as a self-driving opportunity play? Or is this more gen EV and ADAS? And where would the ramp in licensing happen? Would it happen within this year?

Rene Anthony Andrada Haas, CEO

Yes. Thanks, Lee, for both questions, giving us a chance to shout out Ethos and Zena here. So Ethos, we haven't announced its future roadmap, but you can imagine it will be something that can continue to get larger than 4 tops. It's not a data center product. It's not the space to think about it. It's going to be much more in an area where it needs to run off a battery and in some cases, batteries that don't get charged very often. So extremely low-power applications, but also small in terms of physical fit and form factor. I imagine something that could physically be on your body that could be doing AI acceleration. So lots and lots of application spaces for that. We're excited about that platform and also more importantly, the long-term potential for it. Zena CSS, and again, thank you for asking about it and also recognizing that we've already got customers signed up for it. The automotive market is an excellent candidate for CSS. The customers in that timeframe don't have decades and decades of SoC experience. They benefit greatly from a design that accelerates time to market. They benefit greatly from the work we do there. There's a lot of growth potential in terms of what we could do with Zena in the automotive sector. The way to think about where it likely lands is in L2 to L4 ADAS type of applications. Working in that control plane and data plane around everything that has to do with the autonomous section. There is increasingly emerging of the IVI subsystem with the ADAS system. Both have huge amounts of software investment and software stacks that run on Arm. It was just natural that we would go with it. We ended up working on it sooner than we thought. When we kicked off CSS, we were purposeful in the markets we entered. This was a market where we were seeing huge benefits. We're very happy to have announced it. We're happy to have a lead customer and the momentum is very strong there. Yes. Thank you for that, and thank you for all the questions. We are at a very unique time in the industry's history. I think we're seeing, obviously, AI being the most transformative technology, innovation, and disruption that will impact our lives going forward. AI requires a huge amount of compute whether you're doing that in the cloud, an automobile, smartphone, a PC, or even small devices like security cameras or earbuds. ARM is the only company on the planet with a compute platform that can address all that. We're seeing the indicators of that now with Grace Blackwell in the data center and customers using our acceleration capabilities inside of CPU with our SME or Ethos products. We are extremely excited about the future, and it could not be a better time to be working in this industry and the growth trajectory of the company, I think, is unparalleled.

Operator, Operator

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