Morgan Stanley forecasts agentic AI will create a $60B CPU market opportunity by 2030.
The artificial intelligence infrastructure landscape is undergoing a fundamental shift as agentic AI systems move beyond single-task generation toward autonomous, multi-step workflows. Morgan Stanley reports that as AI systems become more complex and require greater coordination, economic value is migrating away from GPUs as the sole bottleneck toward a broader infrastructure stack where CPUs and memory are emerging as critical components. "Agentic AI widens the trade beyond GPUs, with CPUs becoming the control plane for multi-step workflows and system orchestration," Morgan Stanley noted. While GPU demand remains robust, the architecture of agentic AI requires more persistent memory, system-level processing, and coordination overhead, fundamentally widening the AI infrastructure spend pool.
Morgan Stanley quantifies this infrastructure opportunity through new demand frameworks. The brokerage estimates $32.5-60 billion of incremental CPU total addressable market by 2030, within a total server CPU TAM of more than $100 billion. CPU-side orchestration can account for 50-90% of total workload latency in agentic systems, materially raising general-purpose compute intensity. On the memory side, agentic workloads could drive 15-45 exabytes of additional DRAM demand by 2030, equivalent to 26-77% of 2027 annual DRAM supply—a structural shift that will reshape data center memory requirements.
The physical architecture of AI clusters is evolving to support this shift. Agentic workloads rely on CPU-centric or hybrid designs to manage reasoning, tool execution, and memory orchestration, pushing CPU-to-GPU ratios higher at the cluster level. Memory is transitioning from passive storage to an active system component supporting persistent context and continuous learning. This architectural evolution drives content growth across CPUs, DRAM, foundry capacity, ABF substrates, and interconnect layers—the full breadth of the semiconductor supply chain.
The investment beneficiaries extend far beyond GPU vendors. Morgan Stanley identifies that "supply-constrained enablers (foundry, ABF, BMC, interconnect) should capture outsized economics as system complexity rises." The winners span CPU vendors, memory suppliers, storage companies, advanced packaging and substrate providers, foundries, equipment makers, and server manufacturers. Morgan Stanley projects the orchestration CPU market could reach $82.5-110 billion by 2030, with agentic workloads driving the bulk of incremental growth.
The transition from generative to agentic AI marks a capital-intensive infrastructure buildout more complex than the first wave of AI scaling. Greater emphasis on low-latency interconnects, high-bandwidth memory, and resilient foundry supply chains will drive premium pricing in already supply-constrained segments. The report suggests that as agentic AI scales over the next five years, the winners will be those delivering end-to-end system efficiency rather than point solutions. "The beneficiaries of this shift are global and full-stack," Morgan Stanley concluded, pointing to a wider and more diversified AI investment landscape where CPUs and memory could rival GPUs as primary drivers of AI-related semiconductor revenue growth.