Intel has shared extensive details about its improved 18A-P process node, an evolution of the baseline 18A that aims to make the company's contract manufacturing more attractive to outside customers. The new node targets future processors and chips for client devices and data centers, delivering higher performance, better energy efficiency, and improved thermal behavior.
Compared with standard 18A, the 18A-P node gives designers a useful trade-off: either a 9% performance uplift at the same power level, or an 18% power reduction while keeping clock speeds and die complexity unchanged. To achieve this, Intel introduced new RibbonFET transistor variants—some optimised for high-performance circuitry, others for power-sensitive zones.
Despite the upgrades, 18A-P stays compatible with existing 18A designs, making it easier to port already developed projects. Still, squeezing out the full benefits will require additional chip-level optimisation. Intel also reduced manufacturing variation significantly: the spread between fast and slow dies shrank by 30%, which should boost yields and increase the share of high-grade silicon per wafer.
Thermal and reliability improvements got their own focus. Intel claims a 50% improvement in thermal conductivity—a big deal for dense, heavily loaded chips. Long-term voltage stability and low-voltage operation have also been refined. Overall, 18A-P comes across not just as a faster 18A, but as a more mature and balanced node that could attract not only Intel’s own products but also large foundry clients like Apple.