What happened
Intel announced during its Q1 2026 earnings call that AI workloads, particularly agentic scenarios, are driving a significant shift in data centre CPU:GPU ratios, potentially reaching 1:1 from the current 1:4-1:8. This demand surge has increased server CPU prices by up to 20% since March and extended lead times to approximately six months. Intel confirmed it is prioritising data centre chip production over consumer CPUs to address unmet demand, which CFO David Zinsner described as "starts with a B".
Why it matters
Infrastructure costs for AI deployments will rise significantly, impacting procurement teams and data centre architects. The shift to a 1:1 CPU:GPU ratio for agentic AI workloads means a substantial increase in CPU requirements per server, directly affecting hardware budgets and scaling capabilities. Extended six-month lead times for server CPUs will delay new AI infrastructure rollouts and upgrades. This follows recent GPU shortages and price increases, indicating a broader compute supply constraint across the AI ecosystem.
