Intel and AMD are rationing server CPUs to Chinese data centers, with Intel warning of lead times stretching to six months for Xeon processors. AMD's delays are shorter but still significant at eight to ten weeks.
The shortages hit a market where stakes are high. China accounts for more than 20% of Intel's revenue and over 25% of global server shipments. Affected customers include Alibaba and Tencent, according to supply chain sources.
The cause: AI infrastructure demand is consuming fab capacity. TSMC is prioritizing AI chips, while Intel faces yield issues on its latest processes. KeyBanc analysts report that 2026 server CPU capacity from both vendors is already sold out to hyperscalers. Intel server CPU prices in China have risen more than 10%, with potential for 10-15% increases overall.
Intel expects inventory levels to bottom out in Q1 2026 before recovering in Q2. AMD, which relies on TSMC for production, projects server CPU growth of at least 50% in 2026 and says it powers more than 50% of workloads at top hyperscalers.
The shortage also reflects memory market dynamics. Customers have been stockpiling CPUs ahead of anticipated DRAM and NAND price increases, compounding supply pressure.
What this means in practice: Chinese enterprises are evaluating domestic alternatives. Hygon's Dhyana processors and Zhaoxin's KH-series chips offer x86 compatibility, though performance lags behind latest-gen Intel and AMD parts. ARM-based server CPUs are gaining traction for specific workloads, but migration from x86 architectures requires significant engineering work.
The real question is whether this is a temporary squeeze or a structural shift. If hyperscaler demand continues to claim the majority of cutting-edge CPU production, enterprise customers may need to fundamentally rethink procurement strategies. Large firms that stockpiled aggressively face no immediate shortages. Others are working through their options.