Intel and AMD are delivering some bad news to their customers in China: the server chips they need are in short supply. The two companies recently warned that they are struggling to keep up with demand, with Intel telling some buyers they might have to wait up to six months to get their orders.
This shortage is a direct result of the global rush to build artificial intelligence. While most of the hype focuses on specialized AI chips made by companies like Nvidia, the standard “brains” of a server—the central processing unit (CPU)—are just as vital. In China, which accounts for about 20% of Intel’s business, the prices for these server chips have already jumped by more than 10%.
Several factors are causing this logjam. First, a new wave of “agentic AI” is taking over. These tools do more than just chat; they handle complex, multi-step tasks on their own, which requires significantly more CPU power than older applications. Second, there is a massive shortage of memory chips. As memory prices began to soar late last year, many companies panicked and tried to buy their CPUs early to lock in lower total costs. This rush simply emptied the shelves faster.
Intel is also dealing with its own internal production struggles. The company has found it difficult to manufacture enough “good” chips per batch, a problem known as a yield issue. AMD faces a different hurdle. It relies on Taiwan’s TSMC to manufacture its designs. However, TSMC is so busy building high-end AI processors for other giants that it has limited space left to produce the CPUs AMD needs.
This situation marks a major shift in the industry. Just a few years ago, Intel owned nearly 90% of the market. Today, that share has dropped to about 60% as AMD gains ground. For Chinese tech giants like Alibaba and Tencent, these delays make expanding their cloud services much harder and more expensive. Intel expects supply to improve later in 2026, but for now, the industry is stuck in a waiting game.
