Key Points

Arm Holdings is positioned to become a leading force in AI inference chips.
Its energy-efficient processor architectures are seeing robust demand, and venturing into chip production opens fresh growth avenues.
Arm's bold earnings-per-share targets for fiscal 2031 signal substantial future gains for the stock.

Artificial intelligence has fueled explosive stock market growth lately, transforming industries through widespread adoption.

Semiconductors stand out among the prime winners from the AI surge. Vast quantities of chips have powered data centers for AI model training in recent years. Yet, inference now claims a growing share of chip usage. This shift makes sense, since inference deploys trained models into action, processing fresh inputs for instant decisions, responses, and suggestions.

Inference workloads are set to surpass training in scale. Here's the reasoning.

Image source: The Motley Fool.

AI Inference Set to Dominate Compute Demand - A Boon for Arm

Training kicks off with vast datasets shaping AI models, while inference activates them in practice. Demand for inference surges as businesses and users embed AI for efficiency boosts.

McKinsey projects inference will eclipse training by 2030, comprising over half of AI data center compute. Further, it could capture 30% to 40% of overall data center compute by decade's end.

Several chipmakers like Intel and Broadcom offer options to tap this trend, crafting specialized inference processors. Intel's server CPUs also serve hyperscalers for these tasks.

British firm Arm Holdings (NASDAQ: ARM), however, may reap the greatest rewards from inference's rise. Long focused on licensing architectures to chipmakers and OEMs, Arm collects upfront fees plus royalties per chip shipped using its IP.

Powerhouses including Apple, Nvidia, and Amazon leverage Arm's IP for mobile, PC, and data center chips.

Arm's edge lies in energy efficiency, ideal for premium smartphones and AI processors stressing performance-per-watt in data centers. Thus, Google, Nvidia, Amazon, Microsoft, and Marvell build AI server CPUs on Arm tech.

Counterpoint Research forecasts Arm dominating server CPUs and custom AI processors, grabbing 90% market share in three years. Now, Arm expands beyond licensing, unveiling its own Arm AGI CPU in March for agentic AI and inference.

Early signs favor investors: CEO Rene Haas noted on the latest earnings call strong reception, with over $2 billion in demand for fiscal 2027-2028 - double initial projections.

Customer response to the Arm AGI CPU has been very strong. We now have more than $2 billion of customer demand across fiscal 2027 and fiscal 2028. This is more than double what we stated at launch.

Announced March 24, the AGI CPU counts Meta Platforms as lead partner. More adopters likely, given pledges of up to $10 billion per gigawatt data center savings and double x86 performance from AMD or Intel.

Arm eyes $15 billion yearly AGI revenue post-five years, fueling stock appreciation.

Arm Shares Poised for Major Long-Term Appreciation via Chip Fabrication

Fiscal 2026 (ended March 31) saw 23% revenue rise to $4.92 billion, powered by licensing and royalties. AGI's projected $15 billion by 2031 could supercharge growth.

Analysts anticipate faster revenue expansion, with AGI impact from fiscal 2028.

Data by YCharts

Beyond three years, data center CPU opportunity hits $100 billion by 2031. Management projects $25 billion total revenue then, with $9.00 EPS - up sharply from 2026's $1.77.

At Nasdaq-100's 35.6x multiple, shares could hit $320, implying 51% upside from now, making Arm a compelling AI play.

Is Arm Holdings Stock a Buy Today?

Arm Holdings offers strong exposure to AI inference growth through its efficient designs and new chip initiatives.