UBS Says Agentic AI Will Drive CPU Demand, With Arm and AMD the Biggest Beneficiaries
Agentic AI can independently break down tasks, call tools and reason iteratively, generating more sustained computing workloads than the one-off interactions typical of generative AI. UBS said data centers will need more CPUs alongside GPUs to handle coordination, scheduling and general-purpose computing. The shift is set to reshape the server-chip market and benefit Arm, AMD and Intel.
UBS's latest report forecasts that the global server CPU market will expand to $170 billion by 2030. Thanks to its power-efficiency advantages, Arm-based chips could capture a 40%–45% market share by then, making Arm a major beneficiary. AMD is also expected to gain from rising demand and increased market share, while Intel could receive support from the overall market's expansion.
All Coverage
1 original reportsThe Backstory
The history behind this eventUBS and TD Cowen Raise Arm Price Targets as Agentic AI Boosts In-House CPU Outlook
Arm has long relied on licensing CPU architectures and collecting royalties, but in recent years it has accelerated its push into developing chips in-house. Because agentic AI requires continuous planning, tool use and task execution, some data-center workloads could shift from GPUs to CPUs, opening a new source of revenue growth for Arm.
UBS and TD Cowen both raised their price targets for Arm on July 20, with the highest target reaching $475. The upgrades reflected optimism that agentic AI will reshape how data-center workloads are divided between CPUs and GPUs. The investment banks estimate that Arm's in-house CPU-related revenue could reach $14 billion by 2030.
Subscribe to Mark Radar Weekly
Every Friday, the week's strongest signals in your inbox. Unsubscribe anytime.
If you search news on Google, you can set Mark Radar as a preferred source—our coverage will show up more often in your results. Set as preferred source on Google →