Global semiconductor leader TSMC has demonstrated impressive progress in the development of the next generation of chips. Over the last three months, the company has made a significant leap in the development of the A14 (1.4nm) process, surpassing the metrics of the previous N2 (2nm) generation at a similar stage of maturity.

Performance and Yield Leap

TSMC CEO C.C. Wei highlighted key metrics during a briefing with investors, confirming the engineers' success. Internal testing of a test component showed that performance reached 90% of the target, while the yield for 256-Mbit SRAM also reached approximately 90%.

Progress was achieved rapidly. As recently as April, the company reported figures of 85% performance and 80% yield. Thus, in just three months, both critical parameters were raised to the 90% level.

Comparison with the N2 Process

To gauge the scale of this achievement, it is worth comparing these figures with the development history of the N2 process. In April 2023, N2 demonstrated 80% of target performance with a yield of 50%. It was only a year later, in April 2024, that these figures rose to 90% and 80%, respectively.

Development of A14 is moving faster. This is explained by the accumulated experience with Gate-All-Around (GAA) transistors, which TSMC first implemented in N2. The manufacturer managed to eliminate many technological barriers, ensuring high structural uniformity and low defect density.

Launch Plans and Market Demand

The launch of mass production of chips using A14 technology is scheduled for the second half of 2028. However, current metrics theoretically allow for mass production to begin earlier if customers are ready. According to C.C. Wei, clients are already striving to complete design work ahead of schedule, which is a positive signal for the market.

Although A14 will not feature the Super Power Rail (SPR) backside power delivery system (which will debut in A12 in 2029), interest in the new process has already been shown by client processor manufacturers, as well as developers of artificial intelligence and high-performance computing (HPC) hardware.

Technical Advantages of the New Generation

The transition to A14 promises a significant boost in performance compared to N2:

  • Performance increase of 10–15% at the same power consumption and transistor count.
  • Power consumption reduction of 25–30% while maintaining frequency and circuit complexity.
  • Transistor density growth of 20% for mixed-signal circuits and 23% for logic circuits.

These improvements make the technology highly sought-after amidst growing demand for chips for smartphones and AI systems.