Last week, Google announced the Pixel 7 series with the new Tensor G2 chipset under the hood. The highly anticipated launch did reveal some details about the in-house chipset, but one flaw was that the chipset was built on Samsung’s 4nm process. Perhaps this is not related to previous rumors that the second generation Tensor chip will be built using a 4nm process. Now, Google has announced that the Tensor G2 actually uses a 5nm process.

According to Android Authority, a Google representative has confirmed that the Tensor G2 is built on a 5nm process. The executive also said that the chip is specifically designed for real-world use cases with improved performance and power efficiency. This approach also allows the US firm to go a step further in machine learning and add new capabilities using next-generation TPU and G2.

So, the Tensor G2 is essentially a 5nm process. However, Google has not yet officially revealed the exact process and manufacturer of the node. But there is no doubt that Samsung is Google’s main partner in this regard. Samsung has two 5nm process nodes on its roadmap: the 5LPE used by the original Tensor and the later 5LPP. It is currently unclear which node process Tensor G2 uses.

But this should come as a surprise as the current process used in flagship processors like the Snapdragon 8 Gen 1 and Exynos 220 is 4nm. The size reduction makes these 4nm processors potentially more powerful than 5nm chips. Next-generation processors are expected to move to the 3nm process, with associated increases in performance and power consumption. Source