IT News on September 28th: On the 24th of this month, user @Alex_Intel_ on platform X shared a picture allegedly from a report by investment bank KeyBanc Capital Markets.
The image indicates that Naga Chandrasekaran, the General Manager of Intel's Foundry Business, expects that the performance gap between two 1.4nm-level advanced logic semiconductor manufacturing process nodes—Intel's Intel 14A and TSMC's A14—will be within 5%. In terms of advanced packaging, even though TSMC introduces CoPoS heterogeneous integration, Intel's Foundry Business will still be able to maintain a leading position thanks to EMIB-T.
The image mentions that Naga Chandrasekaran reversed the declining trend of Intel's foundry business through accountability and cultural changes. Intel's foundry still plans to achieve profitability by the end of 2027, which roughly requires external customer orders for 40,000 wafers per month.
The image also states that the Intel 18A has a production capacity of nearly 30,000 wafers per month, with performance falling between that of TSMC's N3 and N2P; the defect density of Intel 18A-P has improved by 7% to 11% on a monthly basis, resulting in a steady increase in yield; Intel 14A has seen rapid growth and can now support a production capacity of 14,000 wafers per month.
On the other hand, Intel's foundry services still need to improve the yield of large-sized chips required by ultra-large-scale customers, enhance the yield of EMIB-T motherboards, and reach an agreement with mobile customers regarding the backside power supply issues on the Intel 14A chip.
