intel prepares low-na fallback for cutting-edge 14a chips
Santa Clara, Saturday, 3 May 2025.
intel is not fully committing to high-na euv for its 14a process node. the chip giant is developing a low-na alternative, ensuring identical performance. this move provides flexibility and reduces reliance on asml’s advanced, and very expensive, high-na euv tools. intel aims to mitigate risks associated with adopting the next-generation lithography. this strategic hedge ensures minimal disruption to customers, regardless of the euv approach ultimately chosen for mass production. intel is reporting yield parity between both solutions.
Strategic move for 14a process
Intel’s move to develop a low-NA alternative reflects a pragmatic approach to managing technological risks and costs [1]. ASML’s Twinscan NXE:5000 High-NA EUV machines cost approximately $400 million each [1]. This considerable expense justifies Intel’s exploration of more cost-effective solutions. By having a low-NA option, Intel can negotiate more favorable terms with ASML or adapt more readily to potential delays or performance issues with High-NA EUV [GPT]. Such strategic flexibility can reassure investors concerned about capital expenditures and supply chain dependencies.
Cost efficiency and market positioning
An Intel representative noted that High-NA EUV reduces costs by eliminating around 40 process steps on certain layers, compared to multi-pass low-NA EUV [1]. However, Intel will only use High-NA EUV on a small number of layers of the 14A node [1]. Intel’s CTO, Dr. Naga Chandrasekaran, stated that the company has data showing yield parity between low-NA and high-NA solutions for both 18A and 14A [1]. This suggests Intel can achieve comparable results with potentially lower capital investment, improving profitability and potentially offering more competitive pricing to foundry customers [GPT].
Technological advancements in 14a
Intel’s 14A process node is slated for a 2027 release and promises a 15–20% increase in performance per watt over the 18A node [2]. The 14A node will introduce ‘Turbo Cell’ technology to boost CPU and GPU performance selectively [2]. This technology could enable chip designers to balance performance, power, and area, especially for AI applications [2]. Multiple customers are evaluating 14A, with a variant (14A-E) planned to include feature extensions [2]. These advancements, coupled with a flexible manufacturing approach, could strengthen Intel’s competitive edge against TSMC [2].
Industry outlook and competitive landscape
TSMC, a key competitor, is not planning to use High-NA EUV for its A14 node [1]. This divergence in strategy highlights different risk assessments and investment priorities. Intel’s dual-track approach could be seen as a more conservative, and therefore, less risky path by investors, especially given past challenges in adopting new process technologies [GPT]. The ability to deliver high-performance chips using either High-NA or low-NA EUV could provide a significant advantage, allowing Intel to adapt to market dynamics and technological advancements more effectively [1][2].