Intel sees future 14A chips close to TSMC’s A14 performance

Intel and TSMC’s own projections point to similar performance gains over their current nodes, leaving the winner between 14A and A14 unclear.

Intel expects its 14A process to perform within 5% of TSMC’s A14, a comparison that does not reveal which company’s node is faster. Intel projects a 15%–20% performance gain for 14A over 18A at equal power, while TSMC forecasts a 10%–15% gain for A14 over N2 on the same basis. Applying those projections to current CPU clock examples, the article estimates a 2%–11.3% possible 14A advantage using Intel’s 5.10 GHz Core Ultra X9 388H and AMD’s 5.00 GHz EPYC 9586F as starting points; using Apple’s 4.78 GHz M6 instead produces a 6.5%–16.2% range. These figures are illustrative calculations, not predictions of future chip frequencies, because they apply process projections to chips with different designs and implementations. The gap between Intel’s “within 5%” assessment and the article’s calculations remains unexplained by the sources.