Arm: Lowering full-year royalty revenue expectations (FY25Q1 conference call minutes)

ARM (ARM.O) released its fiscal year 2025 first quarter report on the morning of August 1, 2024, after the US stock market closed.

The following is a summary of the Arm 2025 first quarter financial report conference call. For the financial report analysis, please refer to " ARM: Guidance Cools, AI Faith Cools "

I. $Arm(ARM.US) Financial Report Highlights:

II. Details of ARM Financial Report Conference Call

2.1. Key Points from Management's Statements:

1) Operational Highlights:

① Revenue: Revenue for this quarter increased by 39% year-on-year, far exceeding expectations.

• License revenue increased by 70% year-on-year, reaching a historical high.
• Royalty revenue increased by 17%, with smartphone royalty revenue growing by 50% year-on-year. Despite achieving only single-digit growth in unit shipments, the adoption rate of v9 technology increased, now accounting for 25% of overall royalty revenue, a 20% increase from the previous quarter.

② Developer Community: Arm has over 20 million software developers, making it the largest developer community in the world. The addition of KleidiAI software library makes it easier for developers to benefit from the Arm computing platform.

③ Computing Subsystem Strategy: Introduced CSS per client, with active CSS collaborations in major markets such as mobile phones, laptops, cloud, and automobiles.

2) Financial Highlights:

① Guidance for the next quarter:

a. Revenue: Expected to be between $780 million and $830 million, with the midpoint indicating flat year-on-year revenue.

• Revenue for the second quarter is expected to be the lowest point of the year, mainly due to timing issues in recognizing license revenue.
• Royalty revenue growth: Expected to accelerate to the low 20% range year-on-year.

b. Bookings: Expected to be one of the highest quarters for bookings in many years in the second quarter.

c. Non-GAAP operating results: Expected to be around $500 million.

d. Non-GAAP earnings per share: Expected to be between $0.23 and $0.27, consistent with previous forecasts.

② Full-year guidance for fiscal year 2025:

a. Revenue: Expected to be between $3.8 billion and $4.1 billion, representing a year-on-year growth of 18% to 27%.

• The midpoint of the full-year royalty revenue growth guidance is in the low 20% range, slightly lower than the previously expected mid-20% range
• The outlook for annual licensing revenue is becoming increasingly optimistic, with expected growth in the low to mid-20% range.

b. ACV Growth: Annual ACV growth is expected to be in the low double-digit range. Licensing, through a combination of backlog orders, renewals, and new licenses, is a leading indicator of future royalty income opportunities. Licenses signed now will generate royalty income over the next 2 to 3 years.

c. Non-GAAP Operating Expenses: Expected to be approximately $2.05 billion, a year-over-year increase of 19%.

d. Non-GAAP Earnings Per Share: Full-year non-GAAP earnings per share guidance is between $1.45 and $1.65.

2.2 Analyst Q&A

Q: We hope you can help us understand the transition from the current licensing advantage to future royalty income growth. Looking back at the IPO process, your advice on licensing has been very successful. When can we expect to see corresponding growth in royalty income? How to correctly evaluate the transition relationship between licensing and royalties over time?

A: We are very excited about the growth in our licensing business. This growth is primarily driven by our continued investment in research and development, especially as the demand for customers to design next-generation SoCs using Arm technology continues to increase, particularly in the AI field, our licensing momentum is strengthening. This is partly because many chips need to run AI models that have not even been invented at the time of chip design.

From licensing to chips entering the market, the entire process may take 3 to 4 years, sometimes even longer. The smartphone industry may be the fastest, taking about 3 years. However, other markets like data centers and automotive may take longer. All this increased licensing activity is a positive signal for future royalty growth. We are licensing more and more v9 technologies, as well as more computing subsystems, which have significantly higher royalty rates compared to previous versions. These are strong predictions for future royalty growth.

Q: Recently launched new products are very eye-catching, including the new Cortex-X925 CPU and Immortalis core. In addition, you have also introduced some lesser-known CPU extensions. I want to know if we are seeing momentum in CPU extensions strengthening? Will there be more such products in the future? Which markets will these extensions serve? How do they promote the mid-term royalty growth you mentioned earlier?

A: We introduce new CPU and GPU product lines every year for the smartphone and PC markets, as well as related architectures. The update cycle for the data center market is approximately once every two years, similar to the automotive industry. This has driven a significant increase in demand for Arm technology.

Regarding the extensions you mentioned, we are fully leveraging the features of the v9 version, especially in security and confidential computing, which are particularly popular in data centers. The customization requirements for AI data centers are growing, as each build requires different hardware components, giving Arm a customization advantage; at the same time, the power requirements of AI data centers are unprecedented, further increasing the demand for Arm technology. Security and confidential computing have become essential requirements for AI data centers

Q: You mentioned some platform authorizations in the data center, and we also saw adjustments in certain areas of the semiconductor industry. In the face of some softness in the industry, what have you observed in the authorization channels?

A: Regarding industry adjustments, inventory sales, and how they affect R&D investments, in the past, they have sometimes impacted future design investments. However, currently, we have not seen any signs of slowing down, especially in the end market. The demand for AI workloads is driving the need for stronger computing power, more CPU cores, and subsystems, requirements that go beyond the basic functions of SoCs and systems themselves. We have not seen any slowdown; in fact, authorization activities may have increased compared to a year ago during the IPO.

From a revenue perspective, all these activities are based on existing Arm platforms. Arm platforms are popular because they support the most software, creating a virtuous cycle: the more software, the more companies tend to develop Arm-based chips.

In summary, although there may be a slight slowdown in end-market demand, we have not seen any slowdown in R&D investments, and competition is relatively low. This makes software developers more inclined to choose Arm as the hardware platform, as almost all hardware is based on Arm.

Q: Smartphone royalties increased by 50% year-on-year, while overall royalties increased by 17%. Can we discuss the situation of royalty increases or decreases in other non-smartphone areas?

A: Firstly, we achieved over 50% growth in smartphone royalty income, which is stronger than unit sales growth. Taking all businesses into consideration, the annual growth rate is around 20%. The personal computer market grew slowly last year, but we expect improvement this year. In the cloud computing market, we achieved over 75% year-on-year growth, the strongest growth we have ever seen. This is mainly due to projects like AWS, Microsoft Cobalt, and Google Axion. We expect this growth momentum to continue to accelerate. In the automotive sector, we achieved around 20% year-on-year growth, mainly driven by rapidly growing segments such as ADAS and IVI. However, in the IoT, networking, and industrial sectors, we observed continued weakness with negative growth. While we expect improvement this quarter, the performance in these areas was indeed poor last year.

Q: The CPU business in data centers is relatively easy to track because deployments are more obvious. However, for parallel processing or accelerator businesses, tracking is not as straightforward. Can you share some information about the market penetration of this segment and how it compares to the CPU business? There are more core accounts in the CPU business, which helps increase the average selling price (ASP). I want to know if this dynamic also applies to the accelerator business?

A: When we discuss AI data centers, especially the combination of accelerators and CPUs, NVIDIA is undoubtedly the market leader. However, Arm's penetration in data centers is gradually strengthening. About a year and a half ago, NVIDIA announced Grace Hopper, based on Arm design, combining Arm CPUs with Hopper GPUsThey recently announced that the Grace-Blackwell platform is about to ship, and we expect its sales to surpass Grace Hopper. With its performance and energy efficiency, Grace-Blackwell is poised to become a leader in the AI data center field.

Currently, most accelerators connected to Arm CPUs are custom-made, and mass production is still in the early stages. However, in the case of Grace Blackwell, it is a chip designed for custom systems, with high complexity and excellent energy efficiency. We expect that other accelerators using Arm will also follow this trend, as the biggest advantage Arm brings to AI data centers is customization ability - being able to customize interconnects, memory, and networking according to requirements while maintaining the highest energy efficiency.

In short, although it is still in the early stages, Arm's growth momentum in AI data centers is strong, and the success of Grace-Blackwell is the best proof.

Q: Regarding smartphone royalties, have you considered the impact of inventory adjustments? What is the outlook for the future? Also, regarding the licensing business, do you maintain the expectation of significant growth in the fourth quarter?

A: Regarding smartphone royalties, we achieved a 50% year-on-year growth rate last quarter, while sales growth was only in the single digits. This year, we expect sales growth to be similar, but due to the transition from v8 to v9, we anticipate that our royalty growth will continue to outpace sales growth. Currently, in the smartphone field, approximately half of the royalty income comes from v9.

The Compute Subsystem is expected to start contributing revenue in the second half of this year, especially in the fourth quarter. Although the initial scale is small, it will become more significant next year. We expect that royalty growth for mobile will significantly outpace sales growth in the coming years, as the growth trajectory of CSS may be similar to the transition from v8 to v9, with an expected maturity period of about 4 years, providing sustained growth momentum.

As for the licensing business, we do expect to see the strongest licensing growth in the fourth quarter. Despite some uncertainty, as we have a large number of transaction channels, it appears that transaction channels are even stronger than 90 days ago. Large licensing transactions typically include renewals and new transactions, with these new transactions usually having a 6 to 9-month cycle. Therefore, we are very confident in the outlook for the licensing business and will provide more precise guidance in the quarterly updates.

Q: In the Chinese smartphone market, have you seen smartphone manufacturers starting to license core or AI core technologies similar to Apple? With the rise in average selling prices, have you observed these manufacturers trying to develop their own cores? If so, when do you expect to see licensing royalty income from these efforts?

A: We do not specifically mention licensing transactions in China, as these are conducted through Arm China. Global macro trends, including the growth in demand for CSS in data centers, automotive industry, cars, and smartphones, also apply in China. The demand in the Chinese market is fundamentally no different from other regions, as China also relies on the global software ecosystem, running iOS, Mac OS, native Android, or Windows systems, all based on the Arm architectureArm China is responsible for revenue pricing in China. We do not mention specific partners individually, but from a macro perspective, market behavior is consistent.

Q: Regarding your comment at Computex about Arm architecture reaching a 50% market share in the PC market within 5 years. Although we have seen Arm-based PCs such as Mac and Chromebook, some supporters of x86 architecture are skeptical, believing that the ability to build excellent AI PCs is unrelated to instruction set architecture. What is your response to this? Can you share more about why this time is different, and why Arm can achieve significant market share growth in the short term?

A: I am glad that there are skeptics about our prediction of occupying 50% market share within 5 years. The core of the issue is why this time is different.

Firstly, multiple operating systems, including Apple's operating system, now have Arm-based solutions, and Apple's comprehensive transition has clearly demonstrated the value they place on it. In the Windows ecosystem, several key factors have changed. Today's Arm products use cutting-edge technology, optimized in collaboration with Microsoft, achieving unprecedented battery life. For example, Dell's XPS series based on Arm products claim to have over 19 hours of battery life, far exceeding competitors. Additionally, early Arm-based Windows PCs used outdated technology designed for phones. Now, these products meet the standards of AI PCs.

To achieve market share growth, we need to expand the supplier base and provide comprehensive coverage from entry-level to high-end. We are confident in the information in the ecosystem and believe these will be achieved in the coming years. With a diverse supplier base, excellent battery life, and uncompromising performance, I see no reason why the Windows ecosystem cannot replicate the success of the Mac ecosystem. Considering that the Mac ecosystem is almost entirely based on Arm, reaching a 50% market share within 5 years is not an exaggeration.

Q: When we see important customers adopting v9 technology in the third quarter and mobile CSS customers starting to grow in the fourth quarter, can you help us understand the model for royalty income in the second half of the year so that we can accurately predict the situation?

A: We are currently facing two main changes: the transition from v8 to v9 and the proliferation of CSS. In addition, there are seasonal factors and industry recovery compared to the same period last year, when the industry was at a low point, and now we are entering a stronger growth cycle.

Regarding royalty growth, we expect the annual growth rate to be around 20%, slightly lower than the previously expected 25%. By quarter, the growth rates from the first quarter to the second quarter, the second quarter to the third quarter, and the third quarter to the fourth quarter are expected to be close to 10% each, achieving an overall growth rate of about 20% for the year.

Q: Since the IPO, the licensing business has fluctuated every quarter, but fortunately, these fluctuations have been positive. What drove the positive growth this quarter? Additionally, after a strong first quarter, do you have any updates on this year's booking expectations?

A: Since last quarter, we have seen some strengthening in our licensing business, which helps offset the slight decrease in our expected royalty income. We expect royalty income to decrease by approximately $75 million, while the growth in licensing revenue is roughly equivalent, with no significant change compared to the previous quarter. We have a good understanding and predictive ability for renewals, especially for Advanced Technology Agreements (ATAs), which are the main drivers of licensing bookings and revenue. Additionally, we have some new transactions based on new technologies, which are more difficult to predict, and the final results will help us determine the specific position of guidance over time, giving us a clearer understanding of these new transactions. As for market demand, partners' demand for Arm designs remains stable and strong, possibly even stronger than 90 days ago. This is a key factor driving our business growth.

Regarding quarterly growth, although we had an exceptional performance last year, you will see a year-on-year decline of about 20% next, as we need to compare with the strong quarter last year. Then in the third quarter, the growth rate will return to near flat, slightly positive, and in the fourth quarter, we will see strong growth again. Looking at the whole year, the growth rate will be in the middle range of 20%, with each quarter's performance being very different, but the average growth rate will be close to around 14% ACV growth. ACV growth is a better long-term growth indicator as it smooths out the impact of revenue recognition, providing a more stable growth view.

Q: The penetration rate of v9 in the overall business has reached 25%, with a higher proportion within the mobile market. How should we expect this higher proportion of v9 royalties to gradually impact business segments beyond mobile phones in the future?

A: In overall business forecasting, you can expect that the v9 royalty mix has increased by approximately 500 basis points, and this trend has been ongoing for three to four quarters. Our forecasts indicate that this growth will be maintained at a similar level, and we will notify you promptly of any changes. In different categories, the mobile sector is significantly ahead in v9 adoption due to the shorter cycle from licensing to product finalization and the presence of early adopters. Infrastructure may be the fastest-growing area for v9 after mobile, followed by automotive, with the Internet of Things (IoT) possibly being the last to follow. Although we make forecasts by category, the 500 basis points growth overall remains the most accurate reflection in our model, and we expect this pattern to continue.

Risk Disclosure and Statement of this Article: Dolphin Research Disclaimer and General Disclosure