Intel Shuffles Production Amid AI Boom but Faces Financial Hurdles

The rise in personal computer prices in the coming months will not only be related to the increased cost of memory. As Intel announced, the company is reallocating the production capacity of its factories from client processors in favor of server Xeons. This was reported by Intel’s Chief Financial Officer David Zinsner during the company’s fourth-quarter 2025 earnings call. According to him, the company seriously miscalculated the demand forecasts for data center products, leading to a capacity shortfall.

Zinsner admitted that six months ago, “every hyperscaler” signaled an intention to cut back on orders for processors with more cores. However, the situation changed dramatically, and demand for Xeon rose significantly in the third and fourth quarters. The Xeon 6 platform is widely used as host processors in GPU systems, including Nvidia DGX B200 and B300, as well as in many configurations based on AMD Instinct accelerators. To meet this demand, Intel began actively reallocating production resources. Zinsner noted that the company is “redirecting everything it can to the data center segment.”

Nevertheless, Intel acknowledges that it does not intend to completely abandon the client business for AI. However, priorities within this segment are noticeably changing. “In the client segment, we are focusing on the mid-to-high price range and are not as concentrated on the lower end. To the extent that we have excess capacity, we are directing it towards data centers,” Intel’s CFO stated. This effectively means that Intel is prioritizing more margin-rich Core series processors, freeing up space for the production of Xeons. As a result, cheap PCs with Intel’s lower-end processors may become less accessible.

Intel is not the only company taking this path. Major memory manufacturers, including Micron, SK Hynix, and Samsung, are also facing capacity shortages amid the AI boom. In recent months, all three companies have begun switching their advanced lines to produce DRAM and high-speed HBM memory for servers and AI accelerators. Against this backdrop, consumer memory prices have more than tripled.

The rise in component prices has become another factor influencing Intel’s strategy. Essentially, the company recognizes that there is no sense in loading factories with client processors if buyers are deterred by high PC prices caused by expensive memory. Meanwhile, Zinsner expects capacity constraints to begin easing in the second quarter of the current fiscal year. The company is counting on improved product yields, as well as the introduction of additional equipment for CPU production using Intel 7, Intel 3, and Intel 18A processes. Partially, the situation should improve with the launch of the Core Ultra 3rd series processors (Panther Lake), which more actively use the 18A process. Currently, this process is applied in only one server product Xeon – Clearwater Forest, which is rarely used in AI systems.

Despite ongoing supply issues, Intel’s CEO, Lip-Bu Tan, characterized the quarter as a step towards business recovery: revenue, gross margin, and earnings per share exceeded internal forecasts. At the same time, Intel is still far from previous performance levels. At the end of the fourth quarter, the company recorded a loss of $591 million, and revenue decreased by 4% year over year, to $13.7 billion. The main source of losses was again the Intel Foundry division – the operating loss amounted to $2.5 billion. Meanwhile, 2025 was noticeably better than the disastrous 2024. By the end of 2025, Intel lost $267 million with revenue of $52.9 billion, whereas a year earlier the loss reached $18.8 billion. For the first quarter of 2026, Intel forecasts revenue in the range of $11.7 to $12.7 billion.