Many MindedIntel Reallocates Client CPU Capacity to Data Center Chips for AI Expansion
Intel has reallocated client CPU production capacity to data center CPUs, accelerating global AI infrastructure growth. The company reported $7.3 billion in revenue for its Data Center and AI business in Q3 2026—a 17% sequential and 80% year-over-year increase. Manufacturing costs for Intel 3 process technology declined by 15% since Q1 2026, while the firm expands Fab 24 in Ireland and accelerates equipment installation at Fab 62 in Arizona for Intel 18A/18A-P chips. Intel also expects additional Intel 3 capacity in Oregon to enter production in the second half of 2027 and approximately 40% additional advanced organic substrate capacity by the same timeframe.
This reallocation directly scales AI infrastructure capacity by prioritizing data center processors over consumer chips. The revenue growth and cost reductions indicate Intel is responding to unprecedented AI demand while optimizing production for high-value applications.
For CONNECTION sector needs, this expands access to affordable AI services by increasing the physical capacity for server infrastructure. However, supply constraints on wafer and advanced organic substrate production specifically for multi-chiplet processors using EMIB/Foveros 3D packaging technologies could limit the speed of adoption for complex AI workloads.
What to watch: Intel’s ability to resolve supply constraints for multi-chiplet processors before the second half of 2027. The source notes these constraints apply only to advanced packaging technologies, not all data center chips.
Source: Tom's Hardware