Intel and Advanced Micro Devices (AMD) are negotiating longer-term purchase commitments with Chinese server manufacturers and data center operators as demand for server processors outpaces supply, highlighting how the artificial intelligence boom is extending beyond AI chips to reshape the broader semiconductor market.
According to two sources cited by Reuters, the U.S. chipmakers are seeking agreements that would lock in customer purchasing volumes for server processors over extended periods, giving manufacturers greater visibility into future demand while helping customers secure access to scarce chips.
The negotiations mark a significant shift in the AI supply chain. While global attention has largely focused on shortages of Nvidia’s AI accelerators, demand has broadened across nearly every critical component used in AI infrastructure, including central processing units (CPUs), high-bandwidth memory, networking equipment, storage systems and advanced packaging technologies.
Register for Tekedia Mini-MBA edition 20 (June 8 – Sept 5, 2026).
Register for Tekedia AI in Business Masterclass.
Join Tekedia Capital Syndicate and co-invest in great global startups.
As AI data center construction accelerates worldwide, suppliers across the semiconductor ecosystem are gaining pricing power and increasingly requiring customers to commit to longer-term purchasing arrangements.
The agreements under discussion generally guarantee purchase volumes rather than fixed prices, allowing Intel and AMD to adjust pricing as market conditions evolve. Most proposed contracts cover roughly one year of supply, although both companies have discussed agreements extending two years or longer with some Chinese customers, one of the sources said.
The move mirrors developments in the memory semiconductor industry, where AI-driven shortages have prompted cloud providers and server manufacturers to sign long-term procurement agreements to secure access to high-bandwidth memory and other critical components.
The emergence of similar arrangements for server CPUs signals that shortages are spreading deeper into the semiconductor supply chain.
Unlike AI graphics processors, which have experienced chronic shortages for more than two years, server CPUs had remained relatively available. That balance is now changing as enterprises deploy increasingly sophisticated AI systems that require not only powerful graphics processors for model training but also large numbers of CPUs to manage servers, storage infrastructure, networking operations and AI inference workloads.
Every AI server depends on CPUs to coordinate system operations, distribute workloads, manage memory and execute many non-parallel computing tasks that GPUs are not designed to perform. As a result, rapid expansion of AI infrastructure is driving simultaneous demand for both processor categories.
The tightening supply has begun pushing prices sharply higher in China.
According to one of the sources, prices for some server CPU products have increased by more than 10% month over month, while certain processors have risen more than 40% since the beginning of the year.
The price increases could raise infrastructure costs for Chinese cloud service providers, internet companies and enterprise customers that are rapidly expanding AI computing capacity. Higher processor costs may also slow deployment schedules as companies compete for limited supplies.
The latest developments follow earlier indications that server CPU availability was becoming constrained. Reuters reported earlier this year that Intel and AMD had informed Chinese customers of extended delivery times for server processors, with lead times for certain Intel Xeon products stretching to as long as six months.
Those supply constraints are expected to be closely watched when Intel reports quarterly earnings on Thursday.
During the company’s earnings call in April, Chief Executive Lip-Bu Tan acknowledged that customer demand continued to exceed available supply, particularly for Intel’s Xeon server processors. Tan told analysts that demand “continues to run ahead of supply,” while also highlighting several long-term commercial agreements signed during the first quarter, including a multi-year supply contract with Google.
The comments suggested Intel has already begun shifting toward longer-term customer commitments as it manages persistent demand for data center processors.
AMD has also pointed to sustained growth in the server market. The company, which is scheduled to report earnings in early August, recently raised its estimate for the global server CPU market to more than $120 billion by 2030, citing expanding demand driven by increasingly autonomous “agentic” AI systems that require significantly greater computing resources.
The outlook underpins expectations that AI infrastructure spending will continue expanding well beyond model training, creating sustained demand for CPUs that support inference, orchestration and enterprise workloads.
China remains one of the world’s largest markets for server processors, supported by aggressive investment in hyperscale data centers, national computing infrastructure and AI clusters. The country’s cloud providers and internet companies continue to expand computing capacity even as U.S. export controls restrict access to the most advanced AI graphics processors from companies such as Nvidia.
Those restrictions have made server CPUs an even more important component of China’s AI infrastructure strategy. While Chinese companies face limitations in acquiring cutting-edge GPUs, they continue to require large volumes of x86 processors from Intel and AMD to build and operate AI data centers.
The negotiations also show that the AI boom is no longer benefiting only manufacturers of specialized accelerators. Demand is increasingly spreading across the entire hardware stack, from processors and memory to networking equipment, power systems and advanced manufacturing capacity.
However, for Intel and AMD, stronger demand for server CPUs provides an opportunity to strengthen long-term customer relationships and improve revenue visibility through multi-year supply agreements. At the same time, it reveals a market where supply remains constrained, and customers are becoming more willing to secure future production through contractual commitments.