- TSMC unveils A14 process technology, advancing AI capabilities with significant improvements over its N2 process.
- A14 set for production in 2028, offering up to 15% speed boost or 30% power reduction, and a 20% logic density increase.
- New technologies showcased at TSMC’s North America Technology Symposium to enhance HPC, smartphone, automotive, and IoT sectors.
TSMC (TSM, Financial) has officially introduced its latest logic process technology, the A14, during its North America Technology Symposium held in Santa Clara, California. This new technology is designed to drive advancements in artificial intelligence (AI), providing up to a 15% speed improvement at the same power level or up to 30% power reduction at the same speed compared to the upcoming N2 process. Additionally, A14 offers over 20% more logic density, marking a significant progression in semiconductor technology.
The A14 process is expected to begin production in 2028. TSMC is evolving its NanoFlex™ architecture to NanoFlex™ Pro, enhancing performance, power efficiency, and design flexibility to meet the increasing demands of AI transformation. TSMC Chairman and CEO Dr. C.C. Wei emphasized the company's commitment to providing customers with innovative solutions that bridge the physical and digital worlds.
Alongside A14, TSMC showcased a range of new technologies, including advanced packaging such as Chip on Wafer on Substrate (CoWoS®) and System-on-Wafer (SoW-X), which are geared towards high-performance computing (HPC) and are slated for volume production in 2027. These innovations aim to support the evolving needs of various industries, including smartphones, automotive, and IoT.
The symposium also highlighted TSMC's N4C RF technology, designed for the next generation of smartphones, offering a 30% reduction in both power and area compared to its predecessor, N6RF+, and supporting future standards like WiFi8. This technology will enter risk production in the first quarter of 2026.
