BEIJING, Feb 6, 2026, 23:25 CST
- Intel has warned some China customers of server CPU delivery delays of up to six months, sources said.
- AMD lead times for some products have stretched to eight to 10 weeks, according to a separate source.
- The squeeze adds to investor jitters after AMD’s recent soft outlook raised fresh doubts about its AI push.
Intel and Advanced Micro Devices have told customers in China to expect longer waits for server central processing units, or CPUs, as demand linked to new artificial intelligence data centers strains supply, people with knowledge of the delays said.
The timing matters because the AI boom is now pinching “plain” server parts, not just the high-end graphics chips used to train and run AI models. For cloud groups and server makers, CPUs and memory still set the pace for what gets built and when.
In China, the people said Intel has flagged lead times of up to six months for some server CPUs, with fourth- and fifth-generation Xeon chips particularly tight. One person said prices for Intel server products in China are up more than 10% in general, though terms vary by contract.
A third source said lead times for some AMD products have been pushed out to eight to 10 weeks. Intel told Reuters rapid AI adoption has driven demand for what it called “traditional compute,” while AMD said it was confident it can meet demand globally, citing supplier agreements and its partnership with Taiwan Semiconductor Manufacturing Co.
The shortages could complicate expansion plans for Chinese cloud providers and big server buyers, including Alibaba and Tencent, the Reuters report said. They also land as Beijing-based AI groups race to add capacity while U.S. export rules keep a close watch on the most advanced accelerators.
The constraints have several roots. Reuters cited Intel’s manufacturing yield challenges, TSMC’s prioritisation of AI chip capacity, and a separate crunch in memory chips that has pushed customers to pull forward CPU purchases to avoid costlier builds later.
Intel and AMD together dominate the server CPU market, but the balance has shifted in recent years. Reuters cited a UBS report saying Intel’s share fell to about 60% in 2025 from over 90% in 2019, while AMD’s rose to more than 20% from around 5%.
For AMD, supply stress adds another pressure point after a bumpy week in markets. The chipmaker’s shares sank 13% on Wednesday after a quarterly revenue forecast that pointed to a modest sequential dip, even with a boost from renewed China-bound AI chip sales.
Bernstein analyst Stacy Rasgon said results were not much beyond “inline” without the China boost, adding near-term AI numbers “are not really inflecting.” CEO Lisa Su said demand for AMD’s next-generation AI servers is set to accelerate sharply in the second half, including shipments to OpenAI and other customers.
Investors have been quick to punish anything short of blowout AI growth. “The expectations for large blowout quarters for AI-related hardware companies have skewed what the market is looking for,” Bob O’Donnell, president of TECHnalysis Research, said after AMD’s outlook. Reuters
But there is a big uncertainty: shortages can fade fast if memory supply loosens or if buyers stretch out build plans, and Intel has said it expects supply to improve in the second quarter. Any shift in U.S. licensing for China-bound AI chips would also change AMD’s revenue mix and customers’ ordering behaviour.
