AMD positioned to evolve from GPU-focused semiconductor supplier to broader AI infrastructure provider as CPU demand rises with agentic AI, investor argues

Data center revenue rose 57% year-over-year to $5.8 billion in Q1; AMD increased its long-term server CPU total addressable market estimate from about $60 billion in 2030 to over $120 billion by 2030.

The shift reflects rising demand for CPUs to handle orchestration tasks (inference coordination, memory movement, context routing, and parallel workload management) as agentic AI (AI systems that autonomously manage tasks) becomes more prevalent. AMD controls both CPU and GPU stacks, positioning it uniquely in this transition.

AMD's Helios platform, combining Instinct GPUs, EPYC Venice CPUs, and Pensando networking into a unified system, marks a potential inflection point as hyperscalers demand integrated, turnkey systems rather than assembling components independently. AMD has begun sampling its next-generation MI450 GPU and plans to ramp production in the second half of 2026.