Late last year, John Pitzer, Intel's Vice President of Corporate Planning and Investor Relations, stated that Intel 18A yields were increasing at a rate of 7% per month. This growth rate aligns with industry expectations for a healthy ramp-up, and Panther Lake is following this trajectory, giving Intel confidence in mass production. Entering 2026, Intel appears to have resolved the early yield issues associated with the Intel 18A process, with yields continuing to rise consistently and predictably.
According to TECHPOWERUP, there is more good news regarding the Intel 18A process. Reports indicate that based on Panther Lake production, the yield rate reached approximately 80% in the second quarter of 2026. This suggests that Intel 18A yields have been steadily improving, and Intel seems to have overcome the yield issues it faced in semiconductor manufacturing over the past few years.
On the other hand, Intel 14A is progressing smoothly. Recently, Intel CEO Lip-Bu Tan stated that the next-generation process will enter production in the first quarter of 2027.
Analysts predict that by the end of 2026, the monthly capacity for the Intel 18A process will reach 40,000 wafers, increasing to 60,000 and 80,000 wafers by the end of 2027 and 2028, respectively. After mass production begins next year, the monthly capacity for the Intel 14A process is expected to reach 6,000 wafers by the end of 2027, further increasing to 24,000 wafers by the end of 2028, excluding collaborations under the TeraFab project.
Previous reports suggested that as Intel 18A yields finally reach reliable levels and Intel 14A advances rapidly, Intel might adjust the dual-foundry production ratio for Nova Lake. Instead of assigning 60% to 70% of compute tile production to TSMC, 80% to 90% could be manufactured by its own foundries.