Intel Admits It Rushed 18A, Says Next-Generation 14A Is Ahead as AI Inference Demand Grows

Intel is seeking to regain ground lost to Nvidia and Advanced Micro Devices after missing the first wave of the artificial-intelligence boom. Chief Executive Lip-Bu Tan is betting on data-center CPUs, custom chips, advanced packaging and foundry services, while the rise of agentic AI is reviving demand for Xeon processors that handle the general-purpose computing needed alongside accelerators. Intel’s planned 14A process is central to its bid to restore manufacturing leadership and attract outside customers.

Intel on July 23 forecast third-quarter revenue of $15.8 billion to $16.8 billion, topping the $15.1 billion average estimate compiled by LSEG. Adjusted earnings are expected to reach 38 cents a share, versus analysts’ projection of 27 cents. Revenue for the second quarter ended June 27 rose 25.4% to $16.13 billion, while adjusted profit of 42 cents a share doubled expectations. Tan said Intel was fully committed to high-volume production using its 14A technology in 2028.

CPUs produced on Intel’s mature process nodes, including Intel 7, have long powered mainstream PCs, servers and industrial equipment. But rising demand for data-center processors driven by generative AI has tightened capacity allocation on older nodes. Intel’s push for its most advanced 18A process is critical not only to supplies for PC brands but also to whether its process transition can gain a firm foothold.

As of July 20, 2026, Intel had urged major PC manufacturers worldwide to expand their use of CPUs made on the 18A process while prioritizing products made on older nodes, including Intel 7, for server and industrial customers. The company also notified some customers that it would no longer accept additional orders for certain older CPUs. Reports did not identify the affected models or disclose the scale or value of the shortages.