Arm Holdings (ARM.US) 2026财年第三季度业绩电话会

文章语言：

简

繁

Minutes

原文

会议摘要

Arm reported a 26% YoY revenue growth to $1.24 billion in Q3 FY26, with royalties up 27% to $737 million, driven by AI and data centers. CEO and CFO emphasized AI, edge, and cloud computing focus, predicting data center revenue to exceed mobile. Arm expects 18% YoY growth next quarter, driven by customer demand and investments in next-gen architectures. The company's strategic direction and strong performance attract significant partner investments, positioning Arm for sustained growth in AI and data center markets.

会议速览

Q3 FY26 Earnings Call: Forward-Looking Statements and Non-GAAP Measures

The third quarter fiscal year 26 earnings call outlines the company's forward-looking forecasts and targets, including potential risks and uncertainties that could impact actual results. Non-GAAP financial measures are discussed, with reconciliations provided in the shareholder letter, emphasizing the importance of understanding projected non-GAAP financial measures.

Arm's Record Revenue Growth Driven by AI and Data Center Expansion

Arm reports a 26% year-on-year revenue increase to $1.24 billion, with royalties up 27% to $737 million, primarily from AI and data center growth. The company anticipates its data center business to surpass mobile licensing, with RD investment increasing and non-GAAP EPS reaching 43 cents, reinforcing the Arm platform's strength and commitment to innovation.

Arm's Compute Platform and CSS: Driving AI Innovation and Royalty Growth

Arm's Compute Platform supports diverse AI workloads, enhancing performance and efficiency. CSS, launched 2.5 years ago, exceeds expectations with 21 licenses, accelerating market entry for customers and boosting Arm's royalty income.

Arm CPUs Pivotal in Scaling Agent-Based AI Inference for Enhanced Data Center Efficiency

The shift towards agent-based AI inference necessitates CPUs with higher core counts and power efficiency, benefiting Arm-based chips. Leading hyperscalers like AWS, Nvidia, Microsoft, and Google are adopting Arm CPUs with increased cores, improving performance per watt. Integrated platform designs, such as AWS Graviton with Nitro DPUs and Nvidia's pairing of GPUs with Arm CPUs, enhance system efficiency, marking Arm's growing centrality in modern data center architecture.

Arm's Pivotal Role in Advancing Edge AI and Data Center Growth

Arm's strategic focus on power efficiency and compute systems is driving growth in edge AI and data center applications. Innovations in Arm-based processors are powering advancements in vehicles, robotics, and AI devices, resulting in record revenues and expanding market presence.

Strong Royalty and Licensing Growth Drives Arm's Financial Performance

Arm reports robust royalty revenue growth, driven by automotive market expansion and strong chip demand. Licensing revenue also surges, fueled by next-gen architecture adoption and strategic customer partnerships, with a notable contribution from SoftBank's AI compute strategy. Non-GAAP operating income and EPS meet guidance, reflecting solid financial health and ongoing R&D investments.

Arm's Financial Guidance and Long-Term Growth Strategy Highlighted

The dialogue outlines Arm's financial guidance, including revenue, royalties, and licensing growth expectations, emphasizing confidence in future revenue and investment in advanced technologies for sustained growth. An upcoming event on March 24 is mentioned, with anticipation for announcements but no pre-event details provided.

AI's Impact on Data Centers and Arm's Role, Plus Royalty Revenue Outlook

A discussion on AI's influence in data centers, highlighting Arm's CPU efficiency for AI agents, and an outlook on royalty revenue growth amid consumer electronics demand risks.

Memory Supply Chain Impact on Smartphone Royalties

Discusses potential 20% volume drop impacting 4% smartphone royalties, mitigated by Cloud AI growth, prioritizing premium market.

SoftBank's Commitment to Arm Holdings Amidst Investment Plans

Speculation arises on SoftBank's potential sale of Arm stock for financing new investments, but the company reaffirms its stance against selling any shares, highlighting confidence in Arm's long-term growth.

Analysis of Royalty Revenue Growth Trends and Impact of Memory Shortages

Discussion focused on the deceleration in royalty revenue growth, attributing it to stronger past performances and current memory shortages. The outlook for next year remains cautiously optimistic, with potential for growth if recent trends continue. Memory and wafer shortages are noted as affecting competitors more than the discussed company.

Clarifying Data Center Revenue and SoftBank's Contribution Impact

A discussion focused on quantifying data center revenue, with an emphasis on SoftBank's contribution running at a 200 million rate, consistent with expectations from previous deals, with no new agreements anticipated.

Data Center Revenue Surpassing Smartphone Business by 2025

The data center revenue, currently in double digits, is projected to grow rapidly, potentially outpacing the smartphone business which accounts for 40%-45% of total revenue, within the next two to three years.

Understanding the Impact of V9 CSS on Smartphone Royalty Rates and Revenue

The dialogue discusses how migrating to V9 CSS in smartphones leads to increased royalty rates annually, offsetting potential revenue loss from reduced unit sales. It highlights that new CSS deliveries correlate with higher royalties, impacting revenue projections by up to 4-6%. The focus is on contractual agreements for royalty increases and their effect on revenue, particularly in the context of V9 CSS for smartphones.

Discussion on Future AI Roadmap and Potential Impact on Fiscal 27

An inquiry regarding the potential development of a custom AI processor by SoftBank and its implications for fiscal year 27 is addressed, with the response noting that no specifics can be disclosed about future products.

Evolving Role of ARM IP in AI Data Centers and Smaller Form Factors

ARM IP's role in AI data centers is shifting, with CPUs taking on more workload traditionally handled by GPUs, especially in AI inference. Innovations are expected in addressing constrained power environments and migrating AI to smaller devices, leveraging ARM's dominance in compute platforms.

CSS Impact on Royalty Mix and Future Projections

Discusses the current impact of CSS on royalty revenue, noting its transition from approaching double digits to being well into the teens. Anticipates CSS could represent upwards of 50% of royalty revenue in the next two to three years, driven by customer demand for shortened cycle times. Highlights customer satisfaction and renewals as indicators of CSS value.

Early Views on Future Year Growth and Guidance

Discussion on exceeding FY26 growth targets and anticipation for FY28 guidance, emphasizing script growth and potential new offerings.

Memory Impact and CSS Adoption in Smartphone Volumes and Royalties

The dialogue discusses the minimal impact of memory on smartphone volumes and royalties, emphasizing seasonality and past events over memory as key factors. It also addresses the acceleration of CSS adoption, attributing it to the value of reduced time-to-market over bill-of-materials costs.

Discussion on R&D Investment Growth and Future Opex Trends in Fiscal Year

The dialogue discusses expectations for R&D expenditure growth relative to revenue, with a note on anticipated moderation post Q1, reflecting strategic planning for fiscal year expenses.

AI's Impact on Software Sector and Chip Design

Discusses the software sector's market trends, AI's role in enhancing business operations, and leveraging AI for innovative chip and system designs.

AI's Impact on Stock Market: Opportunities and Uncertainties in the Tech Sector

Discusses AI's role in technology, emphasizing the ongoing need for hardware despite AI advancements. Highlights early stages of AI integration in enterprises, noting potential for growth amidst market uncertainty. Acknowledges significant investments in compute demand, signaling Arm's promising future.

Exploring AI's Impact on Memory Technologies and Power Efficiency in Computing

Discussion highlights the importance of addressing power efficiency and memory technologies like SRAM in AI-driven computing, emphasizing ongoing research and innovation to meet increasing demands in constrained environments.

Arm's Record Quarter Highlights Long-Term Growth in AI and Compute Platforms

Arm delivered its best quarter in history, showcasing strength in royalties and customer shifts. The company anticipates significant opportunities in AI sectors—physical, cloud, and edge AI—aiming to dominate as the preferred compute platform. Future quarters are expected to be even more promising, reflecting Arm's bullish outlook and strategic focus on innovation and market leadership.

要点回答

Q：What is the projected future growth of the data center business compared to mobile licensing?

A：The data center business is expected to grow and become larger than the mobile licensing business in a few years.

Q：How is the Arm platform uniquely suited for AI deployment across various devices and environments?

A：The Arm platform is uniquely suited for AI deployment due to its ability to deliver high performance, energy efficiency, and flexibility across a broad range of power envelopes and use cases, supporting AI workloads from milliwatts to gigawatts.

Q：What is the significance of Compute Subsystems (CSS) for Arm and its licensing progress?

A：Compute Subsystems (CSS) are significant as they help customers lower integration risk and complexity, allowing them to get to market faster. To date, 21 CSS licenses have been signed across 12 companies, with five customers shipping CSS-based chips, including two with second-generation platforms. This creates a tailwind to royalties as demand scales.

Q：How is the shift towards inference reshaping data center design and how does it benefit Arm?

A：The shift towards inference is reshaping data center design by making it increasingly agent-based, which benefits Arm because Arm-based CPU chips deliver industry-leading performance per watt, enabling customers to scale core counts and operate within power, cost, and performance constraints.

Q：What are some examples of leading hyperscalers launching new products with increased core counts?

A：Some examples include AWS launching its fifth-generation Graviton processor with 192 cores, doubling from Graviton 4 and delivering better performance; Nvidia's next-generation Barra CPU featuring 88 Arm-based cores; Microsoft's Cobalt 200 built on the higher-performance Arm-based neoverse N3 platform; and Google previewing its second Arm-based server processor with Thorium powered N4A instances with up to 2x better price performance.

Q：How are integrated platform designs improving system efficiency and enabling more AI output?

A：Integrated platform designs are improving system efficiency by often translating to more AI output or more tokens per watt within the same power envelope. Examples include AWS integrating Graviton with Nitro DPs and training accelerators, and Nvidia pairing GPUs with Arm-based BRCPs and Arm-based Bluefield DPs.

Q：What growth opportunities are opening up in the edge and physical AI markets?

A：In the edge and physical AI markets, there are growth opportunities due to real-time constraints and the need for efficient, power, and latency-sensitive operations. Systems that continually monitor inputs and invoke models to preserve battery life are essential, and the common software foundation across devices, vehicles, and robotics enables scaling deployments without rebuilding software stacks.

Q：How is Arm positioned in the transition of AI to everyday devices?

A：Arm is positioned to support the transition of AI to everyday devices by providing strength in power efficiency, latency, and always-on operation. With a developer ecosystem of over 22 million developers, more than 80% of the global total, Arm is facilitating the shift of AI to every environment and power envelope.

Q：How did the automotive market contribute to the company's royalty performance?

A：The automotive market contributed to the company's strong royalty performance by growing double digits year on year.

Q：What is the significance of the new licensing agreements mentioned?

A：The new licensing agreements, particularly with leading smartphone handset Oems, are significant as they reflect the continued investment by customers in the company's next generation Arm technology.

Q：What are the company's expectations for revenue growth in the upcoming fiscal year?

A：The company expects revenue growth of about 18% year on year, with royalties up low teens year on year and licensing up high teens year on year. They anticipate non GAAP operating expenses to be approximately $745 million and non GAAP EPS to be 58 cents.

Q：How is the company planning to sustain growth in the long term?

A：The company plans to sustain long-term growth by investing in next generation architectures, compute subsystems, and silicon that enable higher performance, greater efficiency, and more AI use cases, leveraging the current strong customer demand and a growing base of long-duration contracts at higher royalty rates.

Q：What is the company's position on the role of CPUs in AI and cloud data centers?

A：The company believes that CPUs are well-suited for the role in AI and cloud data centers due to their power efficiency, always-on capability, and fast latency. They have observed increased deployment of CPUs to address specific workloads related to AI agents and workflows, which is positively contributing to the company's revenue.

Q：How might potential demand destruction in consumer electronics impact the company's royalty revenue?

A：The company has analyzed potential demand destruction in consumer electronics and believes that even a significant reduction in volume would have a limited impact on their royalty revenue. Specifically, a 20% reduction in volumes could lead to a worst-case scenario of a 4-5% impact on smartphone royalties, which would result in a potential 1-2% negative impact on total royalties across the business.

Q：What are the expectations for growth in the Cloud AI or infrastructure business?

A：The Cloud AI or infrastructure business is expected to continue growing ahead of expectations and is more than compensating for risks in the memories and mobile sectors.

Q：What is the position of SoftBank regarding the sale of Nvidia stock?

A：SoftBank is not interested in selling any shares of Nvidia stock, as Masa, the representative from SoftBank, is very bullish on the company's long-term prospects and has no interest in selling.

Q：What factors could impact the forecasted deceleration in royalty revenue growth?

A：The forecasted deceleration in royalty revenue growth could be due to more difficult sales comparisons, particularly from the strong performance in the previous quarter, and the memory shortage affecting growth rates.

Q：What is the expected impact of memory shortages on royalty revenue growth next year?

A：The impact of memory shortages on royalty revenue growth next year is expected to be minimal, with the overall magnitude of royalties expected to be close to earlier forecasts, although the exact impact will depend on whether recent strength continues.

Q：What were the actual and expected contributions from SoftBank regarding data center revenue?

A：The contribution from SoftBank regarding data center revenue was initially estimated to be around $178 million but turned out to be $200 million due to a full quarter's impact. The expectation for data center revenue moving forward is a run rate of $200 million.

Q：How is data center revenue expected to grow and compare to the smartphone business?

A：Data center revenue is anticipated to grow significantly, potentially reaching the teens to near-20% growth over the next two to three years, and could become as large as, or even larger than, the smartphone business which contributes 40% to 45% of total revenue.

Q：What effect will the migration to V9 have on smartphone units and royalty revenue?

A：The migration to V9 is expected to result in a shift to new CSS for smartphones with increased royalty rates. This transition should offset lower smartphone units and result in a maximum 4 to 6% revenue impact within smartphones, assuming the higher rate per unit already contractually agreed upon and scheduled to ship later in the year.

Q：Should one expect an Arm custom AI chip from Nvidia given the partnership with SoftBank?

A：While there is a substantial partnership with SoftBank and a significant revenue from the partnership, there are no specifics or commitments regarding any potential Arm custom AI chip from Nvidia.

Q：How should one think about the penetration rate of Nvidia IP in AI data centers and its evolution?

A：The evolution of Nvidia IP penetration in AI data centers and the exact rate are not specified, but the focus is on how this metric evolves over time, which remains to be seen.

Q：What is the anticipated evolution of data center chip architecture?

A：The anticipated evolution of data center chip architecture involves a morphing of the workloads traditionally handled by CPUs and GPUs. CPUs are expected to increasingly perform the work previously done by GPUs, with more specialized CPU-based custom chips emerging due to the rise of AI inference workloads.

Q：What role does AI play in the expected evolution of computing hardware?

A：AI is expected to be a driving force in the evolution of computing hardware, with AI workloads anticipated to run on every piece of hardware with compute capabilities. This will significantly influence the direction of computing hardware, especially given that the vast majority of current compute platforms are based on the Arm architecture.

Q：What percentage of the royalty mix is CSS currently, and what could it become in the next 2 to 3 years?

A：CSS has already had a material impact on the company's revenue, with the percentage of the royalty mix from CSS currently in the double digits and expected to rise into the teens. Over the next couple of years, it is anticipated that CSS could contribute upwards of 50% of the royalty revenue, driven primarily by the need for customers to shorten their cycle times.

Q：What is the expected growth rate for the upcoming fiscal year, and can early views on the following year be provided?

A：The expected growth rate for the current fiscal year (Fye 26) is at least 20%, which has been exceeded as the guidance is now for 22% at the midpoint. For the next fiscal year (Fye 27), the company expects a reasonable growth rate, though it has not provided a specific percentage. For the subsequent year (Fye 28), while the company has not offered an explicit view, it suggests staying tuned for future updates as opportunities are contemplated, and potential impacts on numbers are still being worked through.

Q：Is memory having an impact on smartphone volumes, and how is the adoption of CSS and V9 expected to be affected by bill of materials challenges?

A：According to the discussion, memory is not having an impact on smartphone volumes as the value gained from accelerating time to market outweighs any potential impacts related to the cost of memory. The adoption of CSS and V9 is not expected to be affected by bill of materials challenges, with the focus remaining on the value created relative to the profits gained by the customers.

Q：What is the anticipated impact of memory on the next quarter's royalties?

A：The impact of memory on the next quarter's royalties is minimal, with the primary driver for the expected growth in royalties being the typical seasonality and Q4 calendar effects. The unusually strong Q4 from the prior year is causing a comparison that is expected to be stronger this year, but the absolute growth in royalties is more influenced by seasonality and lapping the unusual timing from the prior year.

有用

Feedback

免责声明：本内容基于第三方人工智能AI算法生成，内容仅供参考。本平台不对其真实性、合法性、准确性负责，不构成任何投资建议，对以上内容不做任何担保与承诺。若有异议请与本平台反馈。