Intel Corporation (INTC.US) 14A has made significant progress: defect reduction rate hits the best level since 22nm.

Intel's 14A (1.4 nanometer) process has recorded a defect density decline that exceeds the company's expectations, achieving the best performance since the 22 nanometer process. CFO David Zinsner stated at the Deutsche Bank 2026 Technology Conference, "We have not seen performance like this since 22 nanometers." The 14A process is planned for risk production in the second half of 2027, with mass production beginning in 2028. With approximately two years remaining before production starts, there have been substantial changes regarding clients: Intel's internal design teams have started developing products on the 14A, and external foundry clients have shifted from evaluating technical data to inquiring about capacity allocation. Benchmarking against 22 nanometers: Good news and three limitations "The defect density of 14A is better than the target curve we set, and the rate of decline in defects surpasses that of any previous node. We have not seen this kind of performance since 22 nanometersand it can be said that 22 nanometers is one of the best processes Intel has ever launched," Zinsner remarked at the conference. The 22 nanometer process was Intels first FinFET technology, launched ahead of the entire industrycompetitors only followed with their 14/16 nanometer nodes in 2014 to 2015. After the Ivy Bridge processors were successfully launched in 2012, 22 nanometers serviced Intels CPU production line until 2016 and continued to be used in other product lines thereafter. However, Zinsners comparison has strict boundaries. He is comparing the defect reduction trajectory of 14A approximately two years before mass production with the performance of 22 nanometers in a similar timeframe around 2010, rather than a direct comparison of the absolute defect levels of the two generations of technology. Additionally, advancements in wafer inspection equipment over the past 16 years have altered the measurement basis for "defects," and defect density itself does not directly equate to yield. The complexity of the 14A process makes this data especially compelling: second-generation Gate-All-Around (GAA) RibbonFET transistors, second-generation back power delivery PowerDirect, and High-NA EUV lithography present technical challenges far beyond those of 22 nanometers. The ability to quickly reduce defect density with such a combination of technologies stands out in Intels recent process historypreviously, the 14 nanometer node was delayed by a year due to insufficient yield, the first-generation 10 nanometer failed, the 20A process was canceled, and defect levels remained high during the mass production of 18A, while Intel 4 and Intel 3 have seen little public progress. Clients: Shifting from "looking at data" to "demanding capacity" Zinsner revealed that Intels internal design teams have started developing products on 14A. He described this group of internal clients as "probably the pickiest of all," stating, "Their choice to design products using 14A is a significant boost to our confidence." The attitudes of external foundry clients are also undergoing a transformative shift. According to Zinsner, CEO Chen Lifeng and his team are currently meeting weekly with clients, noting, "Clients have shifted from looking at data to asking, 'How much capacity can I get? How should the supply cadence be arranged?' We now have conviction regarding our external clients for 14A." There are still about two years until mass production of 14A begins. Intel has indicated that this node is expected to become a long-lasting process, just like the 22 nanometer node. This article is reproduced from "Wall Street Watch," edited by GMTEight: Feng Qiuyi.