Artificial intelligence (AI) has evolved significantly from its origins as simple chatbots, which typically provided a single response to user queries. Today’s AI models are substantially more advanced, capable of managing complex, multi-step objectives. These capabilities involve collecting necessary data, utilizing advanced models to break down goals into smaller steps, and incorporating external tools such as search engines and databases. This has given rise to autonomous or semi-autonomous AI agents that can plan and execute tasks using software tools.
Market Dynamics and Server CPU Demand
The increasing functionality of AI agents is projected to drive exponential market growth. Industry estimates suggest that the market for agentic AI could expand at a compound annual growth rate (CAGR) of 40% over the next seven years, potentially generating $206 billion in revenue by 2033. This rapid expansion is fueling a significant increase in demand for server central processing units (CPUs).
CPUs are critical components in supporting agentic AI, as they help manage the complex, multi-step workflows by coordinating the required hardware and software resources. This technical shift is changing the typical ratio of CPUs to GPUs (graphics processing units) in AI servers, which is moving from an earlier 1:4 ratio toward 1:2 or even 1:1. Experts suggest this indicates a potential 2-4x increase in demand for server CPUs.
Current Industry Leaders: AMD and Intel
This shift has benefited established semiconductor manufacturers like Advanced Micro Devices (AMD) and Intel. AMD reported that its data center revenue experienced a 107% year-over-year growth in the second quarter of 2026. During the recent earnings call, AMD CEO Lisa Su stated:
In server, we delivered our fifth consecutive quarter of record server CPU revenue with cloud and enterprise sales each growing more than 70% year-over-year, exceeding the outlook we provided last quarter.
Similarly, Intel reported a strong 59% year-over-year increase in its data center and AI (DCAI) business in Q2, reaching $6.3 billion. The company anticipates sustaining this growth with its current generation of server CPUs, known as Granite Rapids. Intel is expanding the production of these chips and collaborating with major original equipment manufacturers (OEMs) including Dell Technologies, Lenovo, and Hewlett Packard Enterprise to bring Granite Rapids to the market.
While both AMD and Intel are expected to benefit from the expanding server CPU market, with AMD forecasting a 50% annual growth rate through 2030, generating $220 billion in revenue by the decade’s end, some analysts suggest a different industry leader may emerge.
The Case for Arm Holdings
British technology company Arm Holdings is noted for providing intellectual property and instruction set architectures (ISAs) that various semiconductor companies utilize when designing chips for smartphones, data centers, and automotive applications. Arm serves as a direct competitor to AMD and Intel by offering Reduced Instruction Set Computing (RISC) architecture, contrasting with the Complex Instruction Set Computing (CISC) x86 platform used by AMD and Intel.
The fundamental difference in architecture is key: while AMD and Intel’s x86 architecture processes complex instructions, Arm’s architecture is recognized for rapidly executing parallel instructions in small batches. Consequently, Arm-based server CPUs are considered ideal for running agentic AI tasks, primarily because they offer between 30% and 60% higher energy efficiency compared to x86 chips.
Market data supports this shift. According to Silicon Analysts, the market share of Arm server CPUs has grown from just 5% in 2020 to approximately 20% currently. Furthermore, market research firm IDC reported in the first quarter of 2026 that revenue from Arm-based server platforms reached $53 billion, significantly surpassing the $34.6 billion revenue recorded by x86 processors.
Arm generates revenue by licensing its IP to customers, such as Nvidia, and by receiving royalties on every chip designed using its architecture. The company has also begun developing its own server CPU, specifically targeting agentic AI and inference workloads. Arm expects to generate over $2 billion in revenue from its server CPUs across the next two fiscal years. The company previously estimated a $1 billion revenue opportunity for its latest product in fiscal 2027 and 2028, but has doubled this estimate due to new customers focused on agentic AI server processors.
Analysts predict that Arm’s revenue growth will accelerate, with an estimated 23% increase in the current fiscal year, reaching $6.06 billion. Moreover, its bottom-line growth is projected to improve, anticipating a 26% increase in ongoing fiscal 2027.
Future Outlook and Market Prediction
According to Counterpoint Research, Arm-based CPUs are expected to capture 90% of the custom server CPU market by 2029, potentially surpassing x86 processors from AMD and Intel. The firm further notes that Arm-based CPUs deliver 2x higher performance-per-watt compared to comparable x86 offerings.
Given its licensing model and superior efficiency profile, Arm is strategically positioned to capitalize on the increasing adoption of Arm-based server CPUs during the growth of the agentic AI era.
