Skip to content
    ARM
    Earnings call· Sep 2025(Q2 FY26)

    ARM HOLDINGS PLC /UK ARM

    Nov 5, 2025 Source

    Executive summary

    Arm Holdings plc Q2 FY26 — Record Revenue and Strong AI-Driven Growth

    Arm Holdings reported a strong second quarter, driven by accelerating demand for AI compute across all markets and the company's energy-efficient platform. Strategic investments in R&D and key partnerships are positioning Arm to capitalize on the growing AI opportunity, from edge devices to hyperscale data centers. The company continues to expand its compute subsystems and architecture offerings, with a focus on delivering performance and efficiency for next-generation AI products.

    Highlights

    5

    • Total revenue reached a record $1.14 billion, up 34% year-on-year, marking the third consecutive billion-dollar quarter.

    • Royalty revenue hit a record $620 million, increasing 21% year-on-year, driven by growth across all major markets.

    • Licensing revenue rose 56% year-on-year to $515 million, reflecting strong demand for next-gen architectures.

    • Non-GAAP EPS of $0.39 exceeded the high end of guidance by $0.06.

    • Data center Neoverse royalties more than doubled year-on-year, with over 1 billion CPUs deployed, and Annualized Contract Value (ACV) grew 28% year-on-year.

    Concerns

    2

    • Power bottleneck for data center expansion

    • Variability of licensing revenue

    Guidance & targets

    6
    CategoryTargetConfidence
    Revenue
    $1.225 billion +/- $50 million
    high materiality
    High
    Royalty revenue growth
    just over 20% year-on-year
    medium materiality
    High
    Licensing revenue growth
    25% to 30% year-on-year
    medium materiality
    High
    Non-GAAP operating expense
    approximately $720 million
    medium materiality
    High
    Non-GAAP EPS
    $0.41 +/- $0.04
    high materiality
    High
    Long-term license revenue growth
    mid- to high single-digit growth
    low materiality
    Medium

    Operational metrics

    9
    Non-GAAP operating expenses
    $648 millionup 31% year-on-year
    Q2 FY26

    Slightly below guidance, reflecting strong R&D investment and ongoing engineering headcount expansion.

    Non-GAAP operating income
    $467 millionup 43% year-on-year
    Q2 FY26

    Resulted in a non-GAAP operating margin of 41.1%.

    Non-GAAP operating margin
    41.1%up from 38.6% a year ago
    Q2 FY26

    Improvement from the prior year.

    Non-GAAP EPS
    $0.39$0.06 above the midpoint of guidance
    Q2 FY26

    Driven by higher revenue and slightly lower OpEx.

    Annualized Contract Value (ACV) growth
    28%year-on-year
    Q2 FY26

    Maintaining strong momentum, exceeding long-term expectations for license revenue.

    Related party revenue (SoftBank)
    $178 millionup $52 million from Q1
    Q2 FY26

    Represents license plus design services for exploring solutions with SoftBank and Stargate partners.

    China sales
    22%
    Q2 FY26

    Demand in China is strong, with licensing being a larger driver of overperformance.

    Developer ecosystem size
    more than 22 million
    Q2 FY26

    A powerful growth engine for Arm, driving software innovation and demand.

    R&D investment for Lumex CSS
    hundreds of millions of dollars
    Past 4 years

    Investment in developing the technology for Lumex Compute Subsystems for smartphones.

    Industry KPIs

    4
    MetricValueDetails
    Ai data center revenuemore than doubledYoY
    Design wins socket pipeline19CSS licenses
    Node platform ramp schedule—
    End market segment revenue mix—

    Product announcements

    4
    ProductTypeDetails
    Lumex CSSlaunch
    Google Pixel 10milestone
    NVIDIA DGX Spark systemlaunch
    Tesla AI5 chipmilestone

    Deals & partnerships

    4
    MetaStrategic partnership to scale AI efficiency across every layer of compute.

    Combines Arm's energy-efficient compute with Meta's AI infrastructure and open technologies to deliver richer, more efficient AI experiences. Focus on consistent compute platform from AI-enabled wearables to AI data centers.

    DreamBig SemiconductorIntention to acquire DreamBig Semiconductor.

    DreamBig has intellectual property around Ethernet and controllers, which are key for scale-up and scale-out networking. The acquisition aims to broaden Arm's offering in high-speed communications for data centers.

    SamsungExpanded collaboration leveraging CSS for its Exynos family of chipsets.

    Samsung is using Arm's Compute Subsystems (CSS) for its chipsets.

    SoftBankPartnership in the Stargate initiative.next number of years

    Stargate is a $500 billion project with OpenAI and SoftBank to build out data centers. Arm is partnering with SoftBank and its partners to provide technology solutions.

    Risks & headwinds

    2
    Power bottleneck for data center expansion

    Access to power has now become the bottleneck

    Mitigation: Arm's compute platform is about 50% more efficient than competitive solutions, addressing the need for efficient compute in power-constrained environments.

    Variability of licensing revenue

    Licensing revenue varies quarter-to-quarter

    Mitigation: Management focuses on Annualized Contract Value (ACV) as a key indicator of underlying licensing trends, which grew 28% year-on-year.

    What to watch in Q3 FY26

    5

    Q4 FY26 Licensing Deals

    Next quarter (Q3 FY26 call)
    CurrentStrong pipeline for remainder of year
    TargetClarity on deals landing in Q4, potential pull forward/pushouts

    Why it matters

    Licensing revenue is variable and a significant component of total revenue; Q4 deals are important for full-year performance.

    But next quarter, we'll definitely have much more clarity around what deals are going to be able to land in Q4 and whether there's any pull forward📎, pushouts or whatnot.

    Q&A highlights

    9

    How is Arm strategically positioned for the increasing demand for AI compute capacity and the build-out of new data centers?

    Rene Haas emphasized that power has become the bottleneck in data center expansion, making Arm's 50% more efficient compute platform highly attractive. He noted that all incremental compute announcements have been Arm-based, driving significant growth in the Neoverse business, which has more than doubled year-over-year.

    “One thing that's become quite evident is that power has become the bottleneck for everyone and power not only means access to energy, but everything underneath it in terms of infrastructure build-out, turbines, transformers, everything associated with generating power. So in that environment, everyone wants to move to the most efficient compute platform as possible. Arm is about 50% more efficient than competitive solutions.”

    asked by Sebastien Cyrus Naji · answered by Rene Haas

    2 min read6 chapters

    Detailed Narrative

    01

    AI Demand and Arm's Strategic Positioning

    Arm is experiencing strong momentum driven by accelerating demand for AI compute, from edge devices to hyperscale data centers. The company positions itself as the sole compute platform delivering AI everywhere, leveraging its energy-efficient architecture. Management highlights that power has become the primary bottleneck in data center build-outs, making Arm's 50% more efficient platform a critical solution for major players like NVIDIA, Amazon, Google, Microsoft, and Tesla.

    02

    Data Center Neoverse Momentum and Stargate Expansion

    The Neoverse compute platform has surpassed 1 billion CPUs deployed, with data center royalties more than doubling year-on-year. This growth is fueled by custom silicon from leading partners, including Google's Arm-based Axion chip, which offers up to 65% better price performance and 60% less energy. The addition of 5 new Stargate sites this quarter further expands visibility into future AI capacity, reinforcing Arm's central role in hyperscale infrastructure development.

    03

    Compute Subsystems (CSS) Success and Lumex Launch

    Demand for Arm's Compute Subsystems (CSS) continues to exceed expectations, with 3 new licenses signed this quarter (smartphone, tablet, data center), bringing the total to 19 CSS licenses across 11 companies. Samsung is leveraging CSS for its Exynos chipsets, achieving up to 40% AI performance uplift. The top 4 Android phone vendors now ship CSS-powered devices. Arm also launched Lumex CSS, its most advanced mobile compute platform, enabling rich on-device AI experiences, with flagship devices from OPPO and vivo expected to ramp later this year.

    04

    Ecosystem Strength and R&D Investment for Future Growth

    Arm's software developer ecosystem now exceeds 22 million, representing over 80% of the world's developer base, acting as a powerful growth engine. The company is accelerating R&D investments to explore opportunities beyond its current platform, including chiplets and complex SoCs, to meet increasing customer demand. This strategic investment aims to capture future growth in compute, ensuring Arm's leadership in the evolving AI landscape.

    05

    Strategic Acquisition and SoftBank Partnership

    Arm announced its intention to acquire DreamBig Semiconductor, a company with intellectual property in Ethernet and controllers for high-speed communications, to broaden its data center offering. The partnership with SoftBank and its Stargate initiative presents a significant opportunity for Arm to provide technology solutions across compute, networking, and power distribution for large-scale data center build-outs, with related party revenue from SoftBank increasing to $178 million this quarter.

    06

    AI at the Edge and Inference Shift

    AI is transforming edge devices, with Google's Pixel 10 featuring the Arm-based Tensor G5 chip for faster and more efficient Gemini model execution. NVIDIA's Arm-based DGX Spark system is shipping for AI developers, and Tesla's next-gen Arm-based AI5 chip delivers 40x faster AI performance for intelligent vehicles. Management anticipates a future shift where inference compute will increasingly move from the cloud to the edge, a domain where Arm's efficient compute is uniquely positioned to excel, supported by partnerships like the one with Meta for scalable software solutions.

    AI-generated summary of the company’s earnings call. Not investment advice.