Market news has been released that NVIDIA will become the first customer of Telco A16 (1.6nm), which can be said to be a significant strategic change, because in the past, NVIDIA has mostly relied on Telco's older processes.
In the past, Apple, Technologies and Qualcomm were the first companies to adopt the latest processes, while NVIDIA focused more on architectural innovation to improve efficiency. However, this is the first technology of NVIDIA to use TECHNOLOGY to advance the A16 technology since 0.11 microns, which is expected to bring huge efficiency improvements and continue the spirit of Morrow's law.
A16 is manufactured to introduce GAAFET architecture and SPR back power supply technology. Among many technologies, power supply on the back is regarded as the most important breakthrough. Market expectation that the time for NVIDIA to go to the A16 process may fall at the end of 2027 or early 2028, as NTD TV regulations begin mass production by the end of 2026.
By the route map, the process may be the first to be applied to Rubin Ultra or Feynman GPUs, which is more likely to be because the Feynman series will also integrate multiple other technological advances.
is also good news for NTD, as almost all mainstream AI hardware manufacturers will adopt advanced NTD processes, and will eventually significantly increase the revenue contribution of this generation of processes.
NVIDIA Reportedly Set to Be TSMC’s First A16 (1.6nm) Customer, Likely Being Influenced By AMD’s Push For High-End AI GPUs Extended reading: Called NVIDIA violated the anti-imperial method! China: Implement further investigations in accordance with the law 2. The nano yield dilemma is resolved? Samsung's mass production before the end of the month is Exynos 2600, expected to be equipped with Galaxy S26