The AI infrastructure boom is driving a surge in demand for central processing units, potentially leading to supply shortages similar to those previously seen with graphics cards. Intel executive Lip-Bu Tan warns that manufacturers may struggle to meet order volumes, with supply constraints potentially lasting until 2028.
The ratio of GPUs to CPUs in AI systems has shifted from eight-to-one to a one-to-one configuration, with some scenarios requiring even more CPUs. This transition is driven by the industry shift from model training to inference, where CPUs are more efficient at handling task coordination and autonomous AI agent operations.
Intel CFO David Zinsner projects that server processor shipments will increase by more than 30% in 2026 and 2027. Lip-Bu Tan estimates that the server CPU market could grow fivefold due to the rise of AI agents. Currently, Intel is already facing difficulties meeting demand for specific server models.
Supply chain challenges are further complicated by rising costs and limited availability of components. Intel reports that memory costs have increased approximately sevenfold. The company continues to face competition from AMD and Arm-based processor manufacturers for contracts with major cloud service providers.
