BEIJING, February 7, 2026, 03:20 (CST)
- Intel has alerted some Chinese customers that certain Xeon server CPUs could face delays of up to six months, according to sources familiar with the notifications.
- A distributor reported that AMD server CPU orders are currently experiencing lead times of eight to 10 weeks.
- These delays pile onto broader supply pressures as data-centre operators ramp up AI infrastructure.
Intel and AMD have alerted their customers in China to expect longer delays for server CPUs, with Intel flagging some shipments could stretch to six months, according to sources familiar with the notices. This bottleneck has pushed Intel server CPU prices in China up by more than 10% in many cases, one source told Reuters.
The warnings highlight that the race to build AI data centres is straining more than just the specialized chips powering AI models. CPUs, or central processing units, handle most server tasks and are found in nearly every data-centre machine.
China makes up over 20% of Intel’s revenue, with major customers like Alibaba and Tencent, sources say. Longer lead times—the delay between ordering and receiving chips—might strain server manufacturers and operators who are already facing higher memory costs.
According to sources, the notices went out over the past few weeks. Intel’s fourth- and fifth-generation Xeon server processors are especially scarce in China, two insiders revealed, leading Intel to limit shipments as demand piles up.
AMD has reportedly alerted clients to supply constraints, according to one source. A distributor handling both CPUs and memory chips noted that lead times for certain AMD server processors have stretched to eight to 10 weeks.
Intel announced that inventories will hit their “lowest level in Q1,” with supply expected to pick up starting in the second quarter. AMD, on the other hand, expressed confidence in fulfilling global customer demand, citing supplier agreements and its collaboration with Taiwan Semiconductor Manufacturing Co.
Intel’s finance chief David Zinsner recently admitted the company misjudged the pace of data-center CPU demand and is now “shifting as much as we can over to the data center.” Meanwhile, AMD CEO Lisa Su expects server CPU demand to remain strong through 2026 and says the company is planning capacity years in advance. Datacenterdynamics
Intel and AMD continue to lead the global server CPU market, despite notable shifts in recent years. Intel’s share dropped to roughly 60% in 2025, down from over 90% back in 2019. Meanwhile, AMD’s slice climbed past 20%, according to a UBS report referenced in Reuters.
The shortage traces back to multiple bottlenecks, according to sources. Intel has hit snags boosting production due to yield problems in manufacturing. Meanwhile, AMD depends on foundry partner TSMC, which is focusing more on AI chip capacity, squeezing the space available for CPUs; yield means the percentage of chips that come off the line fully functional.
Memory chip shortages — another essential server part — have also squeezed supply. The distributor noted that as memory prices started climbing in China late last year, customers rushed to buy CPUs early, aiming to secure cheaper memory costs.
Sources say demand for agentic AI is adding pressure. These systems, built to handle multi-step tasks rather than simple chatbot replies, need more CPU power than typical workloads.
Both chipmakers anticipate supply easing later this year, but this quarter looks tight. If AI data-centre spending surges or memory prices climb again, server buyers in China might face longer wait times and higher costs.
