China is facing an emerging constraint in server central processing units that could complicate the next phase of artificial intelligence expansion. Industry sources say both Intel and AMD have recently notified customers of lengthening delivery times for key server CPUs, highlighting that shortages are no longer confined to AI accelerators alone. For a market that represents more than one-fifth of Intel’s global revenue, the implications extend well beyond near-term pricing pressure.

Lead Times Stretch as Demand Intensifies

According to people familiar with the situation, Intel has warned Chinese clients that delivery lead times for certain fourth- and fifth-generation Xeon processors could extend to as long as six months. AMD customers, meanwhile, have been told to expect waits of eight to ten weeks for some products. These timelines mark a sharp shift from historically predictable server CPU supply and reflect how rapidly demand has intensified.

Prices have responded accordingly. Intel’s server CPU prices in China have reportedly risen by more than 10% on average, although contract structures mean the impact varies by customer. For cloud providers and server manufacturers, this volatility complicates cost planning at a time when AI-related capital expenditure is already elevated.

AI Infrastructure Is Pulling in the Entire Stack

The shortages underscore a broader structural trend. Booming investment in AI infrastructure has created intense demand not only for specialized chips but also for traditional compute and memory. As AI workloads grow more complex, CPUs remain essential for orchestration, data preprocessing, and increasingly for so-called agentic AI systems that perform multi-step operations.

Memory constraints have amplified the problem. With memory prices rising sharply, some Chinese customers accelerated CPU purchases to lock in system configurations, effectively front-loading demand. That behavior reflects a familiar psychological dynamic in supply-constrained markets: fear of future scarcity drives buying decisions that worsen short-term shortages.

Market Structure and Manufacturing Constraints

The supply squeeze is also shaped by how the CPU market is structured. Intel and AMD together dominate global server CPUs, but their manufacturing models differ. Intel has faced persistent challenges in ramping production efficiently, while AMD relies on external manufacturing through TSMC. TSMC’s prioritization of AI chips has left less capacity available for CPUs, tightening supply further.

Market share trends add another layer. Intel’s share of the server CPU market has fallen from above 90% in 2019 to around 60% in 2025, while AMD’s has climbed past 20%, according to a January report from UBS. In China, major customers include cloud giants such as Alibaba and Tencent, whose scale magnifies the impact of any disruption.

Strategic and Market Implications

Intel has acknowledged CPU supply constraints publicly, saying rapid AI adoption has driven strong demand for “traditional compute” and that inventories are expected to bottom in the first quarter before improving through 2026. AMD has emphasized confidence in its supply chain and partnerships, but lead-time extensions suggest near-term pressure remains.

For investors and policymakers in both the U.S. and Israel, the episode is a reminder that AI-driven growth is stressing the entire semiconductor ecosystem. CPUs, long considered mature and predictable, are now part of the strategic chokepoint.

What to Watch Next

Looking ahead, the key variables will be manufacturing yields, foundry allocation decisions, and whether demand normalizes as procurement cycles adjust. If CPU shortages persist, they could slow AI deployment timelines and reinforce the premium placed on supply-chain control. The next phase of the AI race may be decided not just by who builds the best chips, but by who can deliver them on time.