.png?w=1024)
Intel and AMD are negotiating longer-term supply agreements with Chinese data center customers as AI infrastructure demand drives server processor prices higher and increases competition for available chips.
According to Reuters, the proposed agreements would generally secure processor purchase volumes for about a year without locking in prices, although some customers are discussing commitments lasting two years or longer. The negotiations reflect growing pressure on the server CPU market as AI deployments require not only graphics processors but also large numbers of CPUs to power storage, networking, and inference workloads.
That broader demand is now reshaping the server CPU market. Reuters reported that some server processor prices in China have climbed more than 40% since the start of the year, with certain products still posting monthly price increases above 10%. The report marks a change for a market that has traditionally avoided the supply shortages seen in AI accelerators and memory chips.
Earlier this year, Reuters also reported that Intel had warned Chinese customers that delivery times for some Xeon server processors could stretch to six months.
Intel CEO Lip-Bu Tan acknowledged the supply pressure during the company's April earnings call, telling analysts that demand "continues to run ahead of supply," particularly for Xeon server processors. Tan also said Intel signed several long-term supply agreements during the first quarter, including a multi-year contract with Google.
AMD, which is expected to report quarterly earnings in early August, has raised its forecast for the global server CPU market to more than $120 billion by 2030, citing growing demand from agentic AI workloads.
China remains a critical market
China remains one of the world's largest buyers of servers as companies and government-backed projects continue expanding AI computing infrastructure. While U.S. export restrictions limit China's access to the most advanced AI GPUs, server CPUs from Intel and AMD remain an important part of AI deployments, increasing competition for available supply.
The reported contracts could help customers secure future processor allocations, but because they do not lock in pricing, buyers could still face rising costs if shortages continue. Reuters said tighter CPU availability may also slow AI infrastructure rollouts for Chinese cloud providers and internet companies.
What this signals for the AI hardware market
The reported agreements suggest the AI infrastructure boom is entering a new phase where CPUs are becoming almost as strategically important as GPUs.
For Intel and AMD, locking customers into longer-term volume commitments provides more predictable demand while allowing flexibility to adjust pricing as market conditions change. For buyers, the tradeoff is less favorable as guaranteed access to processors comes without protection against future price increases.
If supply remains constrained, other regions could eventually see similar purchasing arrangements, extending the AI hardware shortage beyond accelerators into nearly every major component needed to build modern data centers.