Arm Holdings PLC

NASDAQ:ARM

Earnings Call Analysis

Arm Reports Record Q4 Revenue and Optimistic FY25 Guidance

Arm ended its fiscal year with a record Q4, reporting a 47% year-on-year revenue increase to $928 million. Licensing revenue surged by 60% and royalty revenue grew by 37%, driven by strong adoption of Arm's v9 technology across smartphones and cloud infrastructure. For Q1 FY25, Arm expects revenue between $875 million and $925 million, a year-on-year increase of 30% to 37%. FY25 revenues are projected to be between $3.8 billion and $4.1 billion, marking a 17% to 27% rise. Arm remains confident in sustained 20% annual growth through FY27, buoyed by robust demand for its IP in AI-driven applications.

Introduction to a Record-Breaking Quarter

This past quarter marked a tremendous end to a fiscal year for the company. Arm reported record revenues, up 47% year-over-year to $928 million. This growth was driven by a 60% increase in licensing revenue and a 37% rise in royalty revenue. The company also achieved a notable non-GAAP operating margin of 42%.

Strategic Drivers and Market Expansion

Arm's expansion strategies, discussed during their IPO roadshow, are now driving significant growth. Notably, the adoption of Arm's v9 architecture has accelerated, contributing to increased royalties. The company is seeing broad adoption across all markets, particularly in smartphones and infrastructure.

AI as a Growth Catalyst

Artificial Intelligence (AI) has been a substantial growth driver for Arm. With an extensive user base and critical role in AI workloads, Arm's licensing activity has surged. This surge is attributed to the growing need for hardware suited for new AI workloads.

Diverse Market Penetration

Arm has made significant strides in various markets. For example, Google announced the Axion processor, a custom Arm chip for data centers. In the automotive sector, Arm debuted its first v9-based autonomous solutions with enhanced features for functional safety. These advancements have diversified Arm's presence across different markets.

Outlook and Guidance

Looking forward, Arm expects to maintain an impressive revenue growth trajectory. For Q1 of the next fiscal year, the company anticipates revenue between $875 million and $925 million, a 30% to 37% year-over-year increase. For the entire fiscal year, the revenue forecast is between $3.8 billion and $4.1 billion, representing a 17% to 27% increase.

Projected Growth in Royalties and Licensing

Arm expects continued strong demand for its IP, with royalties projected to grow in the mid-20% range annually. This growth is fueled by the adoption of Arm's v9 architecture and increased market share in cloud and automotive sectors.

Financial Projections

For the next fiscal year, Arm projects non-GAAP operating expenses to rise by 19% to $2.05 billion, reflecting ongoing investment in R&D. The company forecasts full-year non-GAAP EPS between $1.45 and $1.65.

Tailwinds from Technology and Strategic Implementation

Arm's technological advancements and strategic initiatives are setting a strong foundation for future growth. The company's licensing revenue, particularly in AI, contributes to a positive financial outlook, with expectations for continued strong performance across various markets.

Earnings Call Transcript 2024-Q4

[Operator]

Good day, and thank you for standing by. Welcome to the Arm Fourth Quarter Fiscal Year Ending 2024 Conference Call. [Operator Instructions] Please be advised that today's conference is being recorded.

I would now like to hand the conference over to your speaker, Ian Thornton, Vice President of Investor Relations. Please go ahead.

[Ian Thornton, executive]

Thank you very much. Good morning and good afternoon, everybody. My name is Ian Thornton, and I'm the Head of Investor Relations at Arm. I would like to welcome everyone to our earnings conference call for the fourth quarter of the fiscal year ending March 31, 2024.

I'm joined today by Rene Haas, the Chief Executive Officer of Arm; and Jason Child, Arm's Chief Financial Officer.

Hopefully, you will all have downloaded and read the shareholder letter. If not, it is available on the Arm Investor Relations website at investors.arm.com. The shareholder letter provides a rich update on our strategic progress in the quarter.

Before we begin, I'd like to remind everyone that during the course of this conference 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 as of today, our actual results are subject to many risks and uncertainties that could cause actual results to differ materially from what we expect. 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 S-1 filed with the SEC on September 14, 2023. Arm assumes no obligation to update any forward-looking statements, which speak only as of the date they are made.

In addition, 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 and a discussion of certain projected non-GAAP financial measures that we are not able to reconcile without unreasonable efforts and supplementary financial information can be found in the shareholder letter that we released earlier today.

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, who has some prepared remarks.

[Rene Haas, executive]

Thank you, Ian, and hello, everyone. So I'm just going to make a few comments to kick off the call and then I'll pass it over to Jason.

But in summary, this quarter, Q4, obviously being the end of our fiscal year, was just outstanding. We have record revenues for this quarter. And for our first fiscal year as a public company being completed, also record revenue, exceeding the high end of the guidance range.

More specifically for Q4, revenue was up 47% year-on-year. Royalty is up 37% year-on-year, and this is really driven by acceleration of v9 adoption, which I'll speak about a little bit more; and also licensing up 60% year-on-year, which is really a function of increased R&D investment to capture the huge opportunity that is all things AI.

Now in looking back, the expansion strategies that we talked about during our roadshow and then IPO are now all driving growth for the company. As mentioned, we had significant royalty growth in the last quarter, up year-on-year 37%, really driven by v9 adoption. And what we're seeing is the acceleration of v8 to v9, which drives not only better royalties but we're also seeing more CPUs inside the chip, which compounds that royalty growth really across all end markets. And the significant driver for that in client has been around smartphones, but broadly, we also see that in our infrastructure business as well. And v9 adoption will only continue to increase.

In the last quarter, we've also seen proof points of our diversification strategy. Google, the latest hyperscaler announced their Axion processor based on Arm custom chip intended for the data center. They chose us largely because of our compute efficiency but also the ability to not only have a high-performing chip but to design an increasingly performant blade, rack and system for a fantastic TCO.

We also announced our very first autonomous solutions based on v9. This is very, very significant as we're now bringing v9 performance to the automotive sector with automotive enhanced features such as functional safety, and we expect huge growth around this area. And we also have introduced the lowest power transformer on the planet, the Ethos-U85 for IoT-based designs.

One of the strategies we put in place that we are most comfortable with in terms of its growth but very confident in terms of its trajectory is around our, what we call, compute subsystems. And these are essentially taking blocks of IP, putting them together into a full solution, verified and validated, that saves customers huge time to market and also gives them a highly performant solution. So we announced our V3 Neoverse CSS this quarter, which will give increased performance and benefits to customers.

The first automotive CSS is now in discussions with our key partners -- to customers in terms of time to market and efficiency. And our first customer in the Neoverse space doing a design, Microsoft, their Cobalt chip is now ramping.

But probably from a more exciting standpoint, we are oversubscribed on this compute subsystem strategy. We have far more demand for the product than anticipated, and we are anticipating growing that significantly over time. Every end market that we approach has a need for CSSs, and we're very excited about talking about them in the future.

All of this is also being driven by AI. What we are seeing is, because Arm has the largest installed base of CPUs on the planet and has over 70% of the world's population using those CPUs, it's natural that, as these AI workloads are now being moved anywhere from the edge devices to the training data center, that they need support from an Arm CPU standpoint. So whether it's from cloud Edge, from GPT to Llama, all AI workloads rely and run on Arm. And we only see this increasing. Our licensing activity is probably the best proxy for that.

The way to think about licensing revenue as it applies to AI is, as software is moving faster than hardware, the hardware designs need to be upgraded quickly to make sure they can capture the needs of these new AI workloads. So because of that, we have seen huge growth in our licensing activity. We talked about that last quarter and it continued this quarter. So based upon this, we are very, very confident of our growth outlook for the upcoming year. This past year was over 20% revenue growth, and we expect that to be even better than that in this year and the upcoming years. Our growth has been accelerating.

Lastly, it's taken Arm 20 years to get to $1 billion in revenue. It took us 10 years to get to $2 billion. This year, we passed $3 billion in only 2 years after our first $2 billion year, and we expect to be near $4 billion this year. The future is very bright, and we'll run in Arm, and I could not be more excited about the future that we have.

And with that, I'll turn it over to Jason.

[Jason Child, executive]

Thank you, Rene.

Q4 was a strong end to a tremendous year for Arm. For the quarter, we grew revenue 47% year-over-year to $928 million, with licensing revenue up 60% and record royalty revenue up 37% while also delivering non-GAAP operating margin of 42%. Also, as well as delivering strong revenues, we also have grown our remaining performance obligations, or RPO, by 45% year-over-year to nearly $2.5 billion. The royalty acceleration and record RPO provides a great setup for FYE '25.

Turning to guidance. I will briefly touch on both the first quarter and fiscal year ending March 31, 2025. For Q1, we expect revenue of between $875 million and $925 million, representing a 30% to 37% year-over-year increase. Non-GAAP operating expense is expected to be $475 million, a sequential decline due to a change in remuneration as we complete our transition away from cash awards to equity awards, and a onetime Q4 expense. Resulting non-GAAP EPS is expected to be between $0.32 and $0.36.

Unpacking Q1 revenue dynamics a little further, we expect royalty revenue to remain strong with year-over-year growth of approximately 20%. This growth is driven by continued Armv9 adoption and recovery in smartphones, along with continued share gains in automotive and at hyperscalers. This is partially offset by weakness in IoT driven by inventory correction in the broader industrial market, as has been widely reported by many of our semiconductor peers. For Q1 licensing and other revenues, we expect slight sequential growth driven by revenue from backlog.

Looking out to fiscal '25, we expect revenues of between $3.8 billion and $4.1 billion, representing a 17% to 27% year-over-year increase. We expect non-GAAP operating expenses of $2.05 billion, representing a 19% year-over-year increase as we continue to invest in R&D to support future growth initiatives. Our full year non-GAAP EPS is expected to be between $1.45 and $1.65.

We believe our license business is best understood by the measure annualized contract value, or ACV. Our outlook for ACV remains strong. We expect low double-digit growth for the year, reflecting the durable demand in Arm's latest IP.

Licensing revenue, however, will continue to be lumpy from period to period due to the timing of revenue recognition. Based on our current forecast, we expect the first half to represent approximately 40% of our license revenue for the year, with Q2 being the smallest and Q4 being the largest quarter of the year. We have high visibility through a combination of backlog renewals and new licenses.

For royalties, we remain very confident in the demand for Arm-based chips and expect full year growth in the mid-20% range driven by continued v9 adoption, market share gains in cloud and automotive as well as chips based on our compute subsystems starting to ramp in the second half of the year.

Looking out further beyond this year, based on the pipeline for new licenses, agreements already signed, and royalty-bearing chips in development or now shipping, we expect to maintain total revenue growth of at least 20% year-over-year for each of the fiscal years ending in 2026 and '27.

I'll now turn the call back over to Ian.

[Ian Thornton, executive]

Thank you, Jason. We will now move to the Q&A portion of the call. [Operator Instructions] I hand back to you, operator.

[Operator]

[Operator Instructions] First question will come from Ross Seymore of Deutsche Bank.

[Ross Seymore, analyst]

Congrats on the strong results and guide. Rene, I had a question on your infrastructure business and specifically the cloud side. You rattled off a number of the disparate design wins that are starting to ramp from a number of the hyperscaler customers. What I was really wondering is, during the IPO process, you talked about 10% market share in that market rising to roughly, I think you said, 27% or so. Are those products that are occurring now and being launched, are they contemplating that original ramp? Or are you seeing an acceleration above and beyond the growth rate that you have laid out for all of us during the IPO process?

[Rene Haas, executive]

I would say I'm still confident in those numbers, Ross. I would say one of the dynamics that has changed a bit since we've said it, and I think a positive for us, is the increased investment in the data center around all things AI. And that bodes well for Arm for a couple of reasons.

First off, when you think about power efficiency and what's required for the data center, the implementation of an Arm-based design gets very, very interesting relative to not only building a custom chip that will be more power efficient but by designing a blade and interconnecting to storage and an overall rack that will be much more power efficient.

Secondly, and I think we chatted about this during the IPO process with Grace Hopper, but I think now with NVIDIA's most recent announcement, Grace Blackwell, you are going to see an acceleration on the data center in these AI applications.

One of the benefits that you get in terms of designing a chip such as Grace Blackwell is, by integrating the Arm CPU with the NVIDIA GPU, you are able to get an interconnect between the CPU and the GPU that allows for a much higher access to memory, which is one of the limiting factors is for training and inference applications. In a conventional system where you might connect to an x86 externally, you have to do that over a PCIe bus, which is much slower. So by using a custom bus in the NVIDIA example, like NVLink, you get much higher memory bandwidth. So I think what that is going to mean is that Arm adoption in the data center will increase probably faster than the numbers that we had indicated, but we're not saying anything official right now.

[Operator]

And our next question will come from Vivek Arya of Bank of America Securities.

[Vivek Arya, analyst]

It's on the v9 conversion. If I look at Slide 15 of the chart presentation that you have, it shows that it took about 7 years or so for v8 to become about half of the base. Rene, how do you expect this v9 conversion? It seems to have gotten off faster off the gates, 20% or so conversion in a year. What's your assumption of what that conversion looks like in your fiscal '25? I think Jason said mid-20% growth or so in royalty. What does that imply for the level of conversion exiting the year?

[Rene Haas, executive]

Yes. I'll let Jason touch on that conversion rate. But what I can say from a high level, what is different between v8 and v9 is, back in the conversion from v7 to v8, we really didn't have an infrastructure business at all to speak about. We do today, and that is all v9. So that is going to be a growth driver that we saw in terms of conversion for v9 from v8 that we just didn't see from v7 to v8.

In the smartphone business, we are seeing a fairly rapid acceleration in the premium handset segment that is moving to v9. And I think it's been pretty well documented that it's the premium handset business that has enjoyed probably better growth than some of the other segments, which has accelerated as well. So those are 2 drivers for v9 that will look, I think, a bit different than v8 and why it should happen more quickly.

I'll let Jason comment on what we think the potential exit rate might be and particularly how it applies to our year-on-year revenue growth projection.

[Jason Child, executive]

Yes. I would say the 500 basis points of growth or as a percentage of total going from 10% to 15% last quarter and then this most recent quarter from 15% to 20%, I think that's a pretty steady clip. It could be plus or minus but that somewhere around 500 basis points of share growth per quarter is probably a reasonable assumption. So that puts you probably out somewhere in the 2- to 3-year time frame that we probably get to the high end, which is probably somewhere maybe around 60% to 70% because there is still going to be some v8 and v9 out there even as we see growth in the v9 adoption.

Keep in mind, I think when we talked about this back at IPO and even last quarter, you do also see an increased royalty that's almost double from v9 when you go to subsystems. And so we start to see our subsystems come online at the end of this year, in the back half of this year, and then that will start to take hold next year. That also probably has somewhere in the kind of 3- to 4-year time frame of becoming a large share of our royalties. So that's the reason why we have high confidence in a royalty forecast that should be pretty strong and kind of approaching or in that 20% range for years to come.

[Operator]

And our next question will be coming from Matt Ramsay of TD Cowen.

[Matthew Ramsay, analyst]

Rene, I wanted to follow up a little bit on the infrastructure business question that Ross sort of started off with. And one of the questions I've been getting from folks is I can see very, very clearly how things like Grace Hopper and Grace Blackwell will be accelerants of pulling Arm into the data center, and you've announced with Amazon and Google and Microsoft other design wins that will do the same thing, on the flip side, in the accelerated data center, if you just look at a percentage of CapEx that goes to CPUs versus accelerators, that accelerator portion is going to be a vast majority, it seems like, and that trend is accelerating pretty rapidly. So maybe you could speak a little bit to Arm's role not just in the head node of accelerated servers but across the accelerator portion of that sort of exciting growth in CapEx.

[Rene Haas, executive]

Yes. Thanks, Matt, for the question. I think if you go back to the original design win we had with the hyperscalers, and that was with AWS and Graviton, one of the benefits they talked about specifically in terms of using Arm was performance per dollar. And the performance per dollar metric becomes increasingly important as you talk about these large CapEx spends having to manage both the CPU and GPU balance.

I think that is candidly also going to be a tailwind for growth with us because not only do you get a better TCO from a chip standpoint in terms of performance per dollar but, again, just launching into the AWS example for a moment. They built a chip called Nitro, which is based on Arm, for storage offload. And when you combine a Nitro SoC with a Graviton SoC into an EC2 compute rack and then you potentially now configure that for acceleration, that's going to give you a pretty compelling cost advantage versus the competition.

So I think candidly, the trade-off, and this is why the question in terms of what's going on with AI and how we think about market share, I think that's going to be a tailwind for Arm because not only with these standard devices can you configure something that's fairly interesting but you can start to imagine, if anyone wants to do a custom implementation, to greatly enhance some of those interconnect features I talked about. It gets very, very interesting. And only Arm allows that flexibility. There's no way to do that with any alternative.

And not only is there no way to do that with the alternative architecture that is the incumbent today but, more importantly, Arm has done all the work with the partners relative to running the boot code, the system, interface, everything to booting the operating system. So we are extremely well positioned to do well in that platform.

So I think the short summary is I think everything you described is actually going to be quite good for us, which is why I think the growth numbers for our market share in the data center were probably better than we had originally articulated.

[Operator]

Our next question will come from Mehdi Hosseini of Susquehanna International Group.

[Mehdi Hosseini, analyst]

Just want to go back to your fiscal year '25 color and specifically on the licensing. Should I assume that the licensing mix rebounded in the second half of fiscal year? Is that going to be driven by more of the smartphone customers renewing for Armv9? Or is that going to be more broad-based?

[Jason Child, executive]

I would expect it to be more broad-based. As I mentioned, it's going to be mostly in Q4. I mean, that's going to be the biggest license quarter of the year. But definitely with the focus on AI and the focus on AI capability in really all the different end markets that we serve, you should expect it to be quite broad-based.

[Mehdi Hosseini, analyst]

And on the smartphone, the royalty will kick in with the smartphone shipment in calendar year '25.

[Jason Child, executive]

Well, our first subsystems will start to launch basically in the last quarter of the year, so you should see it in our Q4, but I guess it would be in calendar Q1 of '25.

[Rene Haas, executive]

Yes. From the time we license the technology to the time we see a royalty, it's probably 2 or 3 years, best case, which is why the confidence level that we have in terms of royalty growth is quite high because those contracts are done, we know the rates, we know the market share, et cetera.

For the new designs, to Matt's question earlier and Ross', if you're trying to develop an SoC that's going to fit in the server and it needs to run Linux and it needs to run SUSE or Kubernetes, et cetera, et cetera, there's really only one choice, Arm. If you're trying to build an SoC to run Windows, there's only one choice, it's Arm. If you're trying to build an SoC to run Android or Gemini, there's only one choice, it's Arm. So the confidence we have in terms of licensing happening is extremely high.

The tricky part is always in terms of the visibility, whether that license is signed on December 15 or January 15, which moves across a quarter boundary. But in terms of the confidence that it's going to happen, it's extremely high because the choices are rather limited if you want to participate in that market. And when you add in factors on the Windows such as Copilot and things such as Gemini for Android, that's why we're seeing an accelerating effort of the licensees to be able to get access to next-generation technology to take advantage of these new AI features.

[Operator]

[Operator Instructions] And our next question comes from Charles Shi of Needham & Company.

[Yu Shi, analyst]

I have a question, maybe a little bit backward-looking. This was a question I got quite frequently over the last quarter about your royalty revenue growth seen in December quarter. And obviously, you see another very strong sequential growth in March.

So in December quarter, since you already disclosed how much of the royalty revenue coming from related parties, which is a proxy of Arm China, we did notice that Arm China actually was the driver for the royalty revenue growth in December, at least on a sequential basis. Outside of China, it's probably only like a 4%-ish of the sequential growth. So wonder if you guys can use this opportunity to clarify, what was driving that over 50% sequential China royalty revenue growth in December quarter? And can you kind of give us a breakdown for the March quarter, which one, either China or the non-China, is driving most of the sequential growth?

[Jason Child, executive]

Sure, this is Jason. I'll take that question. So first, I would say there has been a little bit of recovery in China in the handset market. It was up, what is it, 1.5%, 2% or so year-on-year going back to the December quarter. So a little bit was just the general market. But the bigger impact was, I would say, the mix shift of the Chinese consumer buying from, I would say, OEMs or partners. Our revenues are based not on where the customer is but based on where the partner who effectively sold the products was. So basically, there was a transfer from Chinese customers buying from Chinese producers versus someone from outside of China. And so what that shows up as, from a royalty perspective, is a shift from rest of world to related parties. And by the way, related parties is mostly China. There are other parties in there as well.

Now in terms of the March quarter end, you'll see similar trends. The rest of the world did accelerate. I mean, we saw a pretty significant royalty acceleration from, what was it, 11%, 12% back in Q3, and it was 37% in Q4. So there is broad-based increase across the world. But you are going to see more acceleration also in China for really the same factors that we saw back in Q3 as well.

[Yu Shi, analyst]

Got it. May I ask a quick clarification, Jason? When you talk about acceleration of the percentages, you mean year-on-year or Q-on-Q?

[Jason Child, executive]

Year-on-year.

[Operator]

Our next question will come from Toshiya Hari of Goldman Sachs.

[Toshiya Hari, analyst]

I had a relatively short-term question on the June quarter outlook for your royalty business, maybe one for Jason. Based on your comments, I think for royalty, you're assuming something like a 7% sequential decline in revenue. Given the v8 to v9 transition, I would think your blended royalty rate continues to grow nicely, so units must be down maybe double digits, again, Q-over-Q. I think typical seasonality is up sequentially if I look at your business over the past couple of years. So I guess the question is, what's driving that sequential decline? I know you're coming off a really high base, but curious what you're assuming, whether it be the smartphone market or some of the other big drivers.

[Jason Child, executive]

Sure. So the driver of the 7% sequential decline is really us looking at the combination of what are we seeing all of our partners forecast either through good faith estimates or what they've actually just disclosed publicly. I mean, you know all the big players, when I look across all those different markets, I'm basically seeing pretty significant declines or weakness, specifically in, I would say, networking and industrial and IoT.

I think on the mobile side, things will be pretty consistent. And so our growth overall will still be very, very much positive, in the kind of 20-ish percent range. But because we're coming off of the kind of the bottom out from over a year ago, the year-on-year will look a little less significant than it did last quarter. But we'll see. If our partners end up seeing different impacts, which, of course, flow through our royalties, there perhaps may be upside. But that's what we're seeing right now, and we'll obviously let you know what we see at the end of the quarter.

[Toshiya Hari, analyst]

And Jason, just to clarify, is it fair to say that automotive and IoT are relative underperformers into June and client and infrastructure should be relatively resilient. Is that a fair statement?

[Jason Child, executive]

Yes, I think it's generally right.

[Operator]

And our next question will come from Ananda Baruah of Loop Capital.

[Ananda Baruah, analyst]

I guess, maybe Rene, I'd love to get your thoughts on PC potential over the next few years and your [ coherence ] in clients given all the work, various folks in the ecosystem that you've been dealing, there seems to be a lot of really interesting reciprocity going on. So just any thoughts on the potential would be great.

[Rene Haas, executive]

Yes, thanks for the question. We've been obviously working on the Windows ecosystem for a long time. The Apple ecosystem has completely converted over. So when we think about our growth, we're talking about Windows. I think over the next number of years, we are very positive about the growth potential.

I think one of the things that's needed for the PC industry to grow, particularly the Windows on Arm segment is going to be a diversification of the supplier base to provide multiple units, multiple SKUs, multiple price points, and multiple experiences for end consumers. Everything I'm hearing says that there are going to be multiple suppliers to serve that market over the next 12 to 36 months. And with that, we think now will be the time, over the next 2, 3 years, where the Arm ecosystem will take a significant level of market share, primarily because of the level of experience that we've seen in the other ecosystem, the fantastic performance, the great battery life, the fact that you can build a high-performance machine minus a fan. I think all those things are going to add up for significant growth. So I think once the vendor base diversifies, I think we're going to see that growth start to kick in over the next 12 to 36 months.

[Operator]

Our next question will come from Gary Mobley of Wells Fargo Securities.

[Gary Mobley, analyst]

When fiscal year '25 is done, you will have grown your licensing revenue 20%, a 20% rate for the prior 2 years, and that's well above what you were predicting during the IPO roadshow. And seemingly, you're predicting long-term 10% growth. So is that the new long-term target we should think about in terms of license revenue growth? And related to that, you highlight how you converted half of your top 30 customers into ATA licensees. How hard will it be to convert the other half?

[Rene Haas, executive]

So thanks for the question. The way I think about the licensing revenue, and you're right, it's a nice upside that continue to be very, very strong, the indications that we get looking forward is that we don't see anything that would stop the licensing activity being a positive momentum. The reason for that is what I mentioned before. If you're trying to enter any of these markets, that are, a, requiring more and more AI; b, require the rich application ecosystem support; and c, broad OS support, the only logical choice that partners have is Arm. And for that reason, we're very confident that we continue to maintain and sustain that level of momentum.

I think in terms of what percentage of customers ultimately move over to an ATA, my estimation is that probably 80% of the customer base at some point in time can be on that, which will give us a lot of increased predictability in terms of license renewals, not so much the if they'll renew but the when, which is [ what we hope for ] going forward.

So Jason, I'll refer to you the second part of that question.

[Jason Child, executive]

Well, let's see, on ATA, I'm not sure if you caught in the notes, we did increase our ATA partners from 27 to 31, so we added 4 more. So a little over roughly half of our, I'd say, at least of the top 20 are there now. But I would say what we're happy about is, when we look into next year even versus what we thought at IPO, we thought that we would probably have closer to 30% year-on-year growth in royalties, and we were all expecting kind of full industry kind of correction. It turns out that industrial IoT and networking aren't quite there. And so even though we are taking down royalty a little bit, we're able to take up license by even more than that royalty reduction mostly because of what Rene just talked about specifically, this demand for additional licensing capabilities very much related to AI.

Now in terms of it being a new normal, I would say no because all license deals, or at least certainly the large ATAs, they have a very high conversion into royalties. Especially with the larger, more kind of mature companies, they typically are pretty good about turning those into design wins and then turning those into royalties. And because these deals are all in v9, they're higher royalty rates than the older counterparts. So I would expect what we talked about back at IPO, when we would get in the next couple of years, get to royalties at 75-plus percent of our total revenue, that's very much still the expectation.

I would say we're just starting to see that, as we get deeper into the v9 penetration life cycle and then also as we see the Compute Subsystems start to come online again at even higher royalty rates at the end of this year, that gives us a great setup for, I would say, a much higher mix of royalties as we go into next year.

[Operator]

Our next question will come from Lee Simpson of Morgan Stanley.

[Lee Simpson, analyst]

I just want to take us back to the v9 profiles that you stood out for automotive, the automotive enhanced profiles, and that includes licensing from Texas, NXP, NVIDIA, et cetera. Just trying to understand how much of that licensing would have been recognized in Q4's number, how much would have gone into RPO. And maybe just related to that, I don't know if I heard correctly but you mentioned multiple customers are working on the CSS release for this in 2025.

[Rene Haas, executive]

Yes. So I'll let Jason go into detail relative to the contribution of the automotive v9 AI technology from a licensing standpoint.

On the CSS engagement, we have multiple partners who are engaged in that. The way to think about the automotive industry is that it is an extremely complex market that needs some degree of customization but also want some level of standardization. And each of the automotive OEMs would love to have a solution that supports a full software stack that would have a number of different contributions relative to differentiation but also something that could be standardized.

So to that level, we've had incredible engagement with lots of different OEMs across this level. And we're very, very confident that the kind of demand that we've seen for CSSs in our other businesses will be there in automotive. It just makes all kinds of sense when you think about the complexity of these devices, the cost involved but yet the need for supply chain resiliency.

So I'll let Jason address the question relative to the licensing revenue and how that all ties together.

[Jason Child, executive]

Yes. So the licensing revenue, in this case, for automotive is a little unique. Typically, we get around 50% of the booking. At the time of booking, it will be booked as revenue. In this case, since v9, it's not quite delivered for auto, it's coming very, very soon, that is actually delayed, so it all is going into backlog. It will be released as soon as it's launched, and we'll announce when that happens.

[Operator]

Our next question will be coming from Timm Schulze-Melander of Redburn Atlantic.

[Timm Schulze-Melander, analyst]

So I just wanted to dig back into the royalty outlook for next quarter and just trying to maybe disaggregate a little bit the volume and the pricing impact. I think you talked about the mix increasing by about 5 percentage points a quarter which, over the course of '25 should give us a pretty decent tailwind. But I think you guided that royalties are going to grow somewhere in the sort of 25% range. Can you maybe just give us a little bit of color about how that breaks down between volume, price, and mix?

[Jason Child, executive]

Okay. So last quarter, if I look at all of our partners, and we don't have all their details on mix, but their growth on average was around 2-ish percent. And obviously, we grew at 37%. So that differential is very much the mix of either more premium chips, higher royalties as well as the higher royalties related to v9. So I would say, going into the next quarter, we're expecting that our overall peer group that we get paid royalties on is probably in that 5-ish percent range. And then we're now expecting that our royalties, as we said, will be somewhere in the 20% range. Obviously, the v9 portion we have a decent handle on, but the mix across the different aspects of the market of premium is just very hard to know ahead of time. So those are the key components.

I would say, going forward, we do expect to see effectively unit growth because, again, I'm assuming that the mix across most of these, we don't get all the detail, but I'm assuming that most of it is going to be pretty consistent. I would assume that we're going to continue to have our royalty growth mostly because of the v9 and mix impacts continue to run pretty significantly ahead of the overall market unit growth.

[Operator]

Our next question will be coming from Harlan Sur of JPMorgan.

[Harlan Sur, analyst]

Backlog was up 45% year-over-year, strong Q4 ACV, so very, very strong end to the year; obviously, a combination of some large renewals but, more importantly, customers. It seems like across all of your end markets, right, adopting the higher value-added compute and Compute Subsystems IP as they sort of look towards their future chip design programs. As you look at your renewal pipeline, discussion with customers and their program timing, looks like you are going to drive strong licensing growth again this year. But does the pipeline suggest that the team can grow total backlog this fiscal year?

[Jason Child, executive]

Harlan, this is Jason. I'll start with that. I'll start with your question. I think at a high level, if I look at last year, you can kind of back into bookings by looking at revenue plus change of backlog. That total number was about $2.2 billion last year. And if you look at kind of what we're assuming for this year, we're definitely assuming less than that. Now last year, we started with a number that was quite a bit below $2.2 billion. And as we mentioned, we had a really strong booking year mostly because of incremental deals that came mostly due to AI. So our assumptions for this year on license is that right now, we'll probably have maybe 60% of the bookings that we had last year, at least that's what it would take to get to kind of our plan.

Now there certainly is opportunity. If you look last year, there was some overage that was probably somewhere in that similar range. We started out with a plan that was probably more like 60% or 70% of what we actually achieved. So we do have good line of sight to the current plan, and we did provide a relatively wide range, mostly because we have to try to factor in the timing of when these deals are going to hit and whether it's Q3 or Q4 or whatever. But if you kind of back up and say, "Well, when I see kind of the midpoint of guidance for this year, what's the confidence level?" I'd say we're at 80-plus percent of the midpoint of our plan. Our guidance is already either in backlog or under contract in royalties.

So really, what we're trying to do is we're trying to forecast what is the incremental bookings that we're going to sign for the year, which a bunch of that's already in the pipeline, and then what possibly could there be beyond that. I'm not really forecasting the stuff that's beyond that, beyond what's in the pipeline today. But if this year looks like last year, that's where some possible upside would come.

[Operator]

I would now like to turn the call back to Rene Haas, CEO, for closing remarks.

[Rene Haas, executive]

Thank you, everyone, and thank you for all the questions, very good set of dialogue and hopefully, we have given some insight in terms of why we are so confident about the future of our company.

This is our third quarter as a public company so it's the third time we've done one of these calls. And it's the third time that we've reported record revenue to you. And it's the third time that we're going to be forecasting growth on the quarter going forward. So we've had 3 quarters of record revenue as a public company. We're also forecasting that next quarter. And I think, again, what you're seeing is the validation of the strategies that we put in place some years ago all coming to fruition, the expansion of our business into multiple markets such as infrastructure, automotive, client PCs, and of course, smartphones. In addition to that, the AI tailwind, which has driven unprecedented growth for our licensing business.

So Arm is a platform company unlike any other. It's a business unlike any other. And the growth and outlook for the company could not have been brighter. To come off of a year where we're talking about 20% growth, and then talking about the year following and the year following when we will do better than that, it's a great business to be in.

So thank you for all your time, and I'll leave it now to Ian or who else to close the call.

[Ian Thornton, executive]

That's all from us, and thank you very much indeed. We'll talk to some of you later and some in 90 days' time.

[Operator]

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