A：The revenue growth in the fourth quarter was driven by a 29% year-over-year growth in licensing revenue, which reached $819 million, and an 11% year-over-year growth in royalty revenue, which amounted to $671 million. The growth was particularly notable in edge AI, physical AI, and cloud AI sectors, with the data center royalty more than doubling year over year.

A：The RAGI CPU, launched at the Arm Everywhere event, is purpose-built for edge AI and is strategically positioned to address the need for more efficient AI data centers. It has the potential to reduce AI data center capital expenditure by up to $100 billion per gigawatt. Key partners like Meta are working on a multi-generation roadmap to support extensive AI infrastructure. This development expands the ways customers can utilize the Arm platform, both through IP compute subsystems and silicon, as well as through the ArmCompute software ecosystem.

A：Customer response to the Arm AGI CPU has been very strong, with more than $2 billion of customer demand across fiscal 2027 and fiscal 2028, more than doubling what was stated at launch. Notable commitments include SAP moving their core database and business application workloads to Arm, starting with AWS Graviton and expanding the use of the RAGI CPU; Cloudflare deploying Arm across its global network; and design wins with key network infrastructure providers like F5 and SK Telecom. Additionally, major cloud providers like AWS, Google, and Nvidia are using Arm-based CPUs as head nodes in their AI systems.

A：Recent customer announcements demonstrate that AI infrastructure is evolving towards a greater integration of Arm-based CPUs. Examples include Google Cloud's TPU 8 t for training and TPU 8 I for inference, which replace x 86 host processors with custom Arm-based Cpu's, resulting in increased performance and power efficiency. Similarly, AWS continues to scale its custom silicon strategy with Arm-based Graviton alongside Turing andNitro, and Microsoft is advancing its Arm-based strategy with Cobalt, designed for high-performance and energy-efficient compute. This indicates that the adoption of Arm-based CPUs is central to performance efficiency and cloud economics.

A：The strategic vision is to grow royalties through IP and CSS and to add silicon as a new growth sector. The focus is on scaling the Arm platform across the next generation of AI workloads. With a strong foundation and continued execution, the aim is to build a future of AI on Arm, extending from cloud infrastructure to the edge and the physical world through a common compute platform and ecosystem. This strategy is underpinned by the momentum in data center AI and the anticipated needs of AI infrastructure in every device, including phones, PCs, vehicles, factories, and sensors.

A：Growth in royalty revenue is primarily driven by the accelerating ramp of Arm based server chips by major hyperscalers and increased deployment of data center networking chips, especially dpu's and sparneck where Arm has significant market share.

A：New strategic partnerships include a long-term agreement with the Indonesian government and two next generation CSS licenses for chip development, one for smartphones and the other for data center networking chips.

A：The amount of license revenue was $819 million, with the SoftBank agreement contributing $200 million.

A：The updated forecast indicates more than $2 billion of demand across fiscal 27 and 28 for the AGI CPU, with the company maintaining its outlook at $1 billion while working on supply chain capacity.

A：By the end of fiscal 31, the projected AGI CPU revenue is $15 billion, and the projected IP revenue is $10 billion, totaling $25 billion, which is expected to translate to more than $9 in EPS.

A：The demand for the new AGI CPU has more than doubled since the last event, and it's a combination of increased forecast from existing customers and new interest from other customers looking to deploy the technology.

A：The company is working with partners to provide different options for the CPU, including finished racks from Supermicro, Lenovo, and other partners, and is securing supply for $2 billion in demand, which includes memory, wafers, packaging, and test equipment.

A：The company is actively managing the increased demand by working around the clock to find solutions and ensure supply availability, while also adjusting guidance accordingly.

A：The company's guidance has not been changed regarding revenue and non-GAAP operating expenses despite the increase in demand.

A：The expected growth for royalty revenues is approximately 28% for the year. This growth is expected to be distributed with a back-end weighting, with around 60% in the second half of the year and 40% in the front half.

A：The expected pattern for Opex growth is an increase of a few percent sequentially every quarter. The overall expense is expected to improve with incremental margins throughout the year, and by the end of the year, the company should be delivering on incremental revenue like it was a couple of years ago.

A：The expectations suggest that while the ratio may not change significantly from a chip standpoint, it may change from a core count standpoint. The growth in the CPU to GPU ratio is not expected to be in the head node of the GPU architecture but rather in the number of CPU cores, which is anticipated to double or quadruple.

A：The potential impact of new CPU-only racks, such as the 256 Vera CPU chips mentioned by Nvidia, could change the CPU to GPU ratio significantly. With more CPU-only racks, the ratio could shift as the demand for CPUs in data centers increases.

A：The shift in business model to include AGI CPU has been communicated to existing IP customers early on, and their support was obtained by explaining the strategy and benefits to the wider ecosystem. The company sought and received support from various partners including EDA partners, software developers, and product licensees. This support was crucial as the customers have products that run on Arm-based chips. The primary reason for the shift was customer demand, and the company is responding to this market demand.

A：Data center royalty revenues are expected to grow significantly in fiscal year 27, doubling year on year as customers build chips based on neoverse.

A：The projected growth rate for license revenue in fiscal 2028 is expected to be more in the 20% range, while long-term targets suggest a range of high single digits to low double digits percent growth.

A：The Opex attributed to supporting the chip business is expected to be more in the 20% range for this year and is embedded in the plan and guidance. Accretion to earnings power may be driven as chip revenues increase and work through existing wafer memory shortage issues.

A：The projected operating margin for the chip business in 2031 is in the 35% percent range, while the IP business is expected to reach a 65% operating margin business or EBITDA margin business. The timeline to reach these numbers will depend on how chip revenue grows.

A：The core count perspective for AGI is to look at it from a core per job basis rather than per token generated by the accelerator. Each agent runs a job independently and does not necessarily require one core per token, suggesting a more complex relationship than a simple core-to-token ratio.

A：The company views larger core counts in CPU chips as more beneficial for power efficiency and the ability to run more batches, leading to increased revenue. This supports the expectation of higher revenue driven by CPU cores in the future.