Intel and AMD are negotiating long-term purchase agreements with Chinese server customers. Driven by AI infrastructure expansion, these deals aim to secure processor volumes for periods of one to two years as prices in China surge by over 40% since the start of the year.
The AI boom has officially moved beyond the GPU. While AI accelerators captured the initial spotlight, the hunger for data center infrastructure has now hit the central processing unit (CPU). Intel and Advanced Micro Devices (AMD) are now pushing Chinese customers toward long-term procurement commitments to manage a market where demand is consistently outstripping supply.
The Shift to a Seller’s Market in China
For years, server CPUs were relatively plentiful compared to the volatile memory and GPU markets. That era is over. The server CPU market is accelerating its transition into a seller’s market, mirroring the path previously taken by memory chips during AI-driven shortages.
This shift is manifesting in aggressive price hikes and extended lead times. Some CPU models in China have seen month-over-month price increases exceeding 10%, with cumulative gains surpassing 40% since the beginning of the year. The supply strain is already creating operational bottlenecks; Intel has warned Chinese clients of delivery windows stretching up to six months for certain Xeon server processors.
Terms of the Long-Term Purchase Agreements
The proposed agreements are designed to guarantee volume rather than price. By locking in the amount of silicon delivered, customers ensure they can actually build their data centers, even if they remain exposed to market price fluctuations. Most of these contracts are expected to span approximately one year, though some commitments have already been extended to two years or more.

This strategic pivot by the U.S. chip giants reflects a broader reality of AI architecture. Large-scale clusters require more than just high-performance GPUs; they need a massive backbone of CPUs to manage storage, networking, and AI inference workloads. As Chinese cloud providers and internet companies race to expand their computing clusters, the CPU has become a critical point of failure in the supply chain.
Intel and AMD’s Strategic Outlook
Intel is already leaning into this model. Tan highlighted that Intel had already secured multiple long-term contracts in the first quarter, including a multi-year deal with Google.

AMD is seeing a similar trajectory, fueled by a specific evolution in AI. AMD has raised its server CPU market forecast to exceed $120 billion by 2030. The company attributes this growth to the rising demand for agentic AI workloads, which place higher demands on general-purpose processing.
Impact on Chinese AI Infrastructure
The convergence of rising costs and longer procurement cycles is creating a direct headwind for China’s digital ambitions. While the U.S. continues to restrict the export of advanced AI GPUs, the demand for the supporting CPU infrastructure remains intense. The result is a double-edged sword for Chinese cloud providers: they are paying significantly more for the hardware they need while facing construction delays for their data centers.
- Price Surge: Some products have climbed over 40% since the start of the year.
- Lead Times: Delivery windows for Intel Xeon processors can reach six months.
- Market Forecast: AMD projects the global server CPU market to surpass $120 billion by 2030.
- Contract Length: Typical agreements cover one year, with some extending beyond two years.
As Intel prepares to report its quarterly earnings, the CPU supply shortage is expected to be a primary focal point for investors. For the Chinese internet giants, the priority has shifted from finding the cheapest chips to simply ensuring they have enough silicon to keep their AI expansion on schedule.
Discover more from Archyworldys
Subscribe to get the latest posts sent to your email.