Google Tensor vs Google Tensor G2

The Google Tensor, codenamed G1, is manufactured using the 5 nm process and is based on the ARM architecture.

It features an 8 MB L2 cache. In benchmarks like Geekbench 6 and Cinebench R23, the Google Tensor shows solid performance in single- and multi-core areas. The integrated graphics unit ARM Mali-G78 MP20 operates with a maximum memory bandwidth of 53 GB/s. Its design is optimized for energy efficiency in mobile applications.

Individual market prices for the processor are not available as it is not sold separately. The Google Tensor G2, codenamed G2, is also manufactured using the 5 nm process and utilizes the ARM architecture.

It has an 8 MB L2 cache and an additional 4 MB L3 cache. In performance tests like Geekbench 6 and Cinebench R23, the Google Tensor G2 exhibits similar performance in single- and multi-core areas. The more modern ARM Mali-G710 MP7 graphics unit enables improved performance in games.

The memory bandwidth is also 53 GB/s. This processor is also designed for high energy efficiency and is not available as a separate product in retail. In gaming scenarios, the Google Tensor G2 performs better thanks to its newer graphics unit. The computing power is comparable, but the Google Tensor G2 offers an optimized overall experience due to its additional L3 cache and more efficient graphics unit.

For improved graphics performance, the Google Tensor G2 is the more recommendable choice; for basic mobile applications, the Google Tensor is still sufficient.