
Today Arm published a letter to its shareholders containing the companyās results for its fiscal third quarter for FYE25 ended Dec. 31, 2024. The infographic below provides the key highlights:

Here are what the results mean and how they were achieved:
- Revenue of $983 million, which is up 19% year-over-year due to record levels of royalty revenue and continued strength in license revenue.
- Royalty revenue of $580 million, which is up 23% year-over-year. This is being primarily driven by the continued adoption of Armv9, the ramp-up of chips based on Arm Compute Subsystems (CSS), IoT improvement and increased usage of Arm-based chips in datacenters.
- License and other revenue of $403 million, which is up 14% year-over-year due to normal fluctuations in the timing and size of multiple high-value license agreements signed during the quarter. Ā
- Non-GAAP operating income of $442 million resulting in a 45% non-GAAP operating margin, and non-GAAP earnings per share of $0.39.
- Arm expects Arm CSS to become the product of choice for many partners to license for future chip designs, with strong demand due to requirements for increasingly complex, customized silicon for specific workloads, and benefits of reduced development costs and time-to-market.
- All major global automakers, including BMW, Honda, Rivian, Nuro and Mercedes rely on Armās energy efficient and safe platform to power automated driving, parking, advanced autonomous capabilities and feature-rich cockpit experiences in their latest vehicles. At CES 2025, a key focus was on the Arm-based NVIDIA DRIVE AGX Thor, the first centralized compute platform for AI and autonomous driving workloads, which is built on Arm’s server-class Neoverse Automotive Enhanced (AE) CPU. The adoption of the same Neoverse platform in the cloud and at the edge ā in the car ā means software developers can develop applications in the cloud and then seamlessly run them in the car, accelerating the time-to-market.
- The emergence of new Arm-based AI hardware ā many of which are devices that do not extensively rely on the cloud for processing and are untethered from the power grid ā and smaller, lighter language models are unlocking edge AI use-cases across consumer technologies, automotive and industrial IoT. Examples include:
- NVIDIAās new Project DIGITS, which combines the Arm-based Grace CPU featuring 20 power-efficient Arm Cortex-X and Arm Cortex-A CPU cores with the Blackwell GPU into the GB10 Superchip to power the worldās smallest AI supercomputer;
- The recently announced Arm-based Samsung Galaxy S25, which brings context-aware multimodal AI agents ā combining voice, video, and text input ā to mobile for the first time; and
- Flagship AI smartphones from OPPO and vivo that have adopted MediaTek’s new Arm-powered Dimensity 9400 system-on-chip (SoC).
- Arm is a key technology supplier to the recently announced Stargate Project, alongside Microsoft, NVIDIA, OpenAI, and Oracle. The news further cements Arm at the very center of one of the largest AI infrastructure investments to date.
- Arm is a strategic partner with OpenAI and the SoftBank Group on āCristal intelligenceā, which enables AI agents that solve complex problems for knowledge work. The Arm compute platform, powered by KleidiAI libraries, is set to provide the performance, efficiency, and scalability required to support the growing computational demands of these new AI agents from edge to cloud.
The full letter to shareholders is available on the Arm investor relations website (https://investors.arm.com/financials/quarterly-annual-results). Information about the Arm earnings conference call can be found here.
Related
Arm Achieves Record Q2 Results
Arm Holdings plc Reports Results for the Third Quarter of the Fiscal Year Ended 2025
Latest on X
šø Spotted at #GTC25 - catch CEO Rene Haas in conversation with Jim Cramer straight from the GTC show floor
š¢ Arm Accuracy Super Resolution is now available to ALL developers!
Now open source under the MIT license & available as a plugin for @UnrealEngine, with a @Unity plugin coming soon, it brings you enhanced visual quality without compromising performance! https://okt.to/e0Blbqā¦
Attending #GDC2025? See you tomorrow at the Arm Developer Summit!
Get ready to create the ultimate mobile gaming experiences with our action-packed agenda:
š WH3020, Moscone Center, San Francisco
The Arm Silicon Startups Contest with Silicon Catalyst is back!Ā šĀ
GetĀ access to ourĀ technologyĀ through Arm Flexible Access for Startups, expert support, and investor connections to fast-track your product to market &Ā accelerate your innovation journey: https://okt.to/PU6Ids
Arm and NVIDIA are redefining enterprise computing with the NVIDIA Graceā¢ CPU & Superchips, bridging efficiency and acceleration.āÆĀ
š Listen how here: https://okt.to/2RtqUy
Embedded World 2025, over and out! š¬
Always a highlight of our annual calendar, it's been great connecting with the AI, IoT and automotive ecosystem to share why the embedded industry is built on Arm.
š· Here's a look at some of our highlights.
Collaboration is key!
Listen as Christopher Rumpf, Senior Director Software and Solutions, Auto, breaks down the importance of the automotive industry working together on SDVs. š¤
š @CipiaVision is improving the driving experience with in-cabin sensing, enabled by AI and computer vision on the Arm compute platform.
By detecting drowsiness, distraction, phone use, seatbelt, driver ID, and more, we're bringing trust and safety to the road.
Here's how itā¦
š How can developers accelerate edge AI?
John Thompson, Senior Director of Software and Ecosystems, IoT, shares how Arm Kleidi for IoT is unlocking CPU performance for the next generation of edge AI applications - with some impressive results so far.
Edge AI is evolving fast. šØ
With the new Armv9 edge AI platform is set to drive the next wave of innovation, Paul Williamson, SVP and GM for IoT, shares his thoughts on what to expect next, from LLMs and SLMs on-device to new use cases across industries.
Listen in.
š
With @alifsemi's latest MCUs and the Arm Ethos-U85, you can efficiently execute transformer networks for genAI workloads.
That means faster, more accurate processing for vision, voice, text and sensor data - without relying on the cloud. āļø
#EmbeddedWorld #ew25
OEMs are increasingly focused on enhancing the emotional connection between humans and their cars.
Christopher Rumpf shares how the SOAFEE SIG makes this possible, experimenting with the cloud-to-car continuum on Arm to shape the future of AI-enabled, software-defined vehicles.ā¦
Here's to the final day of Embedded World 2025!
We're ready for a jam-packed day filled with insight, conversations and innovation.
One stars of the show has been our SoCks!Ā š§¦
If you picked up a pair - past or present - weād love to see them in action. Share your best SoCksā¦
Tech leaders take on āThis or Thatā at Embedded World 2025! š„
We put some of Armās brightest minds in the hot seat. Watch now and play along!
#ew25
šØ Don't miss Rene Haas and @LizClaman today at #SXSW 2025 for an inside look at the state of AI, Armās role in it, and what to expect next!
š 2:30ā3:30 PM CT
https://okt.to/Y9AxlB
Spotted at Embedded World! š
Have you managed to catch any of our speakers at the show?
š Generative AI is enabling rapid advancement in the automotive industry, and we are right at the front of it.
Watch below as Suraj Gajendra, VP of Products and Solutions for Automotive, explains the impact of AI, and what new user experiences are coming into play.
How are we supporting the growth of IoT development for our ecosystem? š
ā”ļøĀ Enhancing AI performance, bringing Arm KleidiAI to the IoT, and investing in the range of devices built on Arm.
Paul Williamson, SVP and GM for IoT,Ā gives us all the details.
Ready to star in your own unique story with our Generative AI Photo Booth? šŖ
Using the YOLO11 model, Raspberry Pi 5 detects objects and creates an amazing story using the TinyStories SLM.
If you're at the event, come check it out, and get it all printed on a postcard to takeā¦
Cortex-A320 supports both rich and real-time operating systems. šŖ
Listen as John Thompson, Senior Director of Software and Ecosystems, talks to us more about it.
š The biggest challenge for vehicle OEMs today?
Ensuring robust software infrastructure, seamless deployment, and planning for the future demands of AI driven applications.
Suraj Gajendra, VP of Products and Solutions for Automotive, breaks it down.ai-d
#EmbeddedWorld #EW25
By leveraging the leading-edge capabilities of the Arm compute platform, @RenesasGlobal's new RZ/V2N brings advanced endpoint vision AI to a huge range of applications, from smart city to industrial.
See it for yourself this Embedded World! #ew25 #EmbeddedWorld

Introducing the new RZ/V2N MPU, targeting high-volume vision AI. With the latest addition of the RZ/V2N, the RZ/V series has now expanded to cover a full range of markets, from the low-end RZ/V2L to the high-end RZ/V2H.
Learn more: https://bit.ly/41VyDx8
#MPU #DRP
Great conversations and valuable connections. š¤ š¬
Earlier this week at #EmbeddedWorld, we brought developers together for an Edge AI roundtable to discuss how we can best address their needs, and the future of IoT Development.
If you joined us, thank you. For those buildingā¦
š¶ Head to the Embedded World booth with us as we kick off day 2 of the show! š
From AI at the edge to cloud-to-car intelligence, explore cutting-edge demos like @AlifSemiās Ethos-U85 for transformers at the edge and @CipiaVisionās in-cabin sensing.
ā Receive a personalizedā¦







