The production technology of the Tensor G6 processor, which is at the center of the Pixel 11 series that Google introduced a few days ago, has become clear. Although the company did not openly share this detail during the promotion, there was a widespread expectation that the Tensor G6 was released from the 2 nm production process. Google Vice President of Hardware Peng Yu-Chun confirmed that the new processor uses TSMC’s 3 nm production process. Thus, after last year’s Tensor G5, Google chose to stay in the 3 nm class in Tensor G6. Despite this, the performance and efficiency values announced by the company show that the processor contains significant improvements in various areas, even if the production geometry does not change.
The expectation that Tensor G6 would be 2 nm arose with the acceleration of the transition to more advanced production processes in the semiconductor industry. However, the nanometer value in which a processor is produced does not alone determine its performance, energy consumption or daily usage experience. Chip design, CPU and GPU architecture, artificial intelligence accelerators, memory system and software optimizations also have a direct impact on the result. Therefore, the fact that the Tensor G6 is produced in the same 3 nm class as the Tensor G5 does not mean that the two processors offer the same features in technical terms. The figures provided by Google also indicate that the company is focusing especially on artificial intelligence computing capacity and energy efficiency this time.
Tensor G6 advances on the performance and efficiency side
According to the information shared by Google, Tensor G6 increases TPU calculation performance by 50 percent compared to Tensor G5. This section, called Tensor Processing Unit, is used to accelerate artificial intelligence and machine learning-based operations on Pixel phones. Therefore, it seems possible that this increase will be reflected in areas such as image processing, artificial intelligence features and machine learning models running on the device. In addition, it is stated that the renewed CPU calculation infrastructure increases power efficiency by up to 30 percent. Higher energy efficiency could impact the daily usage performance of the Pixel 11 family in terms of keeping battery drain and temperatures under control, especially under heavy processing loads.
However, the percentages announced by Google are not enough to reveal the performance of Tensor G6 in real usage conditions. It will be possible to see more clearly in which workloads the up to 30 percent higher power efficiency statement is achieved and to what extent this is reflected in the battery life of the Pixel 11 models through independent tests. Similarly, how much time the 50 percent calculation increase on the TPU side saves in users’ frequently used features will vary depending on the application. Moreover, the temperature and frequency management of the processor under long-term high performance can be as decisive as the values in the technical specifications table. Therefore, when evaluating the Tensor G6, the performance, battery life and temperature tests of the Pixel 11 series will need to be considered together, rather than the production process.
On the other hand, even if the Tensor G6 was actually produced with 2 nm technology, it would not be the first 2 nm mobile processor. Samsung had already acted in this area by announcing the Exynos 2600 in December 2025. Although the transition to smaller processes in semiconductor production generally provides the potential to increase transistor density and reduce power consumption, the results achieved by manufacturers may differ depending on the architecture used and design preferences. For this reason, it may be misleading to evaluate expressions such as 2 nm and 3 nm as a direct ranking of performance between phones. In the Tensor G6 example, it seems that Google’s focus is on improving TPU capacity and CPU efficiency on the existing 3 nm infrastructure rather than minimizing the production process.
The decisive point for the Pixel 11 family will be to what extent Tensor G6 can carry the announced improvements into daily use. Google’s own Tensor processors have been used in past Pixel generations not just for raw performance but also to more tightly integrate photo processing and AI-based features into the device. While the 50 percent higher TPU computing capacity of Tensor G6 continues this approach, the CPU power efficiency increase of up to 30 percent stands out as a more concrete improvement on the battery consumption side. On the other hand, preserving the 3 nm production process cannot be considered an advantage or disadvantage of the processor alone. Comprehensive performance, battery and temperature measurements of the Pixel 11 series will show more clearly how technical changes in Tensor G6 are reflected in the user experience.
Join Channel