Investors turn to AMD, Arm, and Marvell as alternatives to Nvidia, citing shift toward inference workloads in AI data centers

Nvidia controls an estimated 80% of the AI chip market but has lagged its chip peers in stock performance this year. AMD, Marvell, and Arm shares have significantly outpaced Nvidia's returns as investors seek exposure to the inference phase of the AI supercycle.

AMD, Marvell, and Arm are designing chips for inference-focused workloads (the step where an AI produces an answer), which Deloitte projects will account for two-thirds of AI computing power in data centers this year. Inference is less compute-intensive than AI training, prompting hyperscalers to adopt CPUs (central processing units) and ASICs (application-specific integrated circuits) instead of GPUs.

AMD expects the server CPU market to grow by 35% annually through 2030, generating more than $120 billion in revenue by the end of that period. Arm anticipates total revenue growing to $25 billion after five years, up from $4.9 billion in the previous fiscal year. Nvidia, AMD, Arm, and Marvell are all projected to clock faster earnings growth than Nvidia in the next two fiscal years according to the earnings table.