NVIDIA has reportedly become the sole customer for TSMC’s next-generation A16 process node, with both companies currently conducting joint testing. The A16 process integrates nanosheet transistors and Super Power Rail (SPR) technology, promising a major leap in performance and efficiency. NVIDIA is expected to adopt the A16 process for its upcoming Feynman GPU architecture, scheduled for release in 2028. Apple, TSMC’s long-time key client, has yet to engage in talks regarding the use of the A16 process for its future mobile SoCs. Apple’s absence is surprising for the industry, and it remains unclear whether the company will follow up. [IThome, in Chinese]
Related
