The first-generation Nintendo Switch has been on the market for nearly 6 years. The processor is a customized chip of NVIDIA Tegra X1, which cannot meet the increasingly complex performance and user experience requirements. There has been news that Nintendo will launch a high-end version of the new machine and start discussions with partners.
Nintendo has said the Switch is doing well and that new opportunities will be announced at the right time. The Switch stiffness exceeded half of the estimated life cycle, which may mean that Nintendo has indeed developed a new machine, but it will not be listed in the near future.
According to market sources, the new Switch chip, code-named Black Knight, is a customized internal code-named Tegra Orin T239 chip by GPU manufacturer NVIDIA. The information shows that the T239 chip has 8 ARM Cortex A78 and A78C cores, and the GPU of the Ampere architecture may include the Lovelace function. The new Switch Graphics API section includes DLSS 2.2 features and support for ray tracing. In terms of foundries, it is expected to be TSMC, which NVIDIA has cooperated with for many years.
NVIDIA Tegra X1 was released in March 2015. It adopts TSMC’s 20nm process, TDP 15W, and the CPU is 4 Cortex-A57 cores with 1.9GHz operation clock and 4 Cortex-A53 cores with 1.3GHz. The GPU is mainly based on NVIDIA’s own graphics card Maxwell architecture, with 256 CUDA cores and a clock pulse of 1GHz.
In addition to the Switch, the Tegra X1-equipped devices include the Nvidia Shield and the Google Pixel C tablet. However, the Tegra X1 of the Switch is an abbreviated version, with 4 A53 cores disabled, and the overall computing clock is only 1.02GHz, which is lower than the 1.9GHz of the full version of the Tegra X1.
(First image source: Flickr/Shinji CC BY 2.0)
New technological knowledge, updated from time to time