Google Tensor G2
Benchmarks & Specs

The Google Tensor G2 is a smartphone processor developed by Google that is used exclusively in the Google Pixel 7 and Google Pixel 7 Pro from Google. The CPU has 8 CPU cores and uses a hybrid core structure. The special feature is that the Google Tensor G2 does not have one very high-clocked Prime core, but two.

The Prime cores use ARMs Cortex-X1 design and clock at up to 2.85 GHz. They are supplemented by two Cortex-A78s that clock at 2.35 GHz. There are also four additional Cortex-A55 energy-saving cores that only work at 1.8 GHz, but are particularly energy efficient. In mobile devices, this can extend battery life because the larger CPU cores are put into a standby state and are only used when they are needed.

When it comes to AI acceleration, the Google Tensor G2 can rely on the Google Edge TPU with an AI computing power of up to 4 TOPS. The Google Edge TPU accelerates, for example, camera functions, image and video processing. AI acceleration can also help ensure that the CPU cores are required as little as possible.

The Google Tensor G2 uses an ARM Mali-G710 MP7 as the graphics card. With a theoretical GPU computing power of 0.7 TFLOPS, this is quite suitable for playing most smartphone games smoothly.

The Google Tensor G2 supports up to 12 GB of DDR5-5500 memory with a maximum memory bandwidth of up to 53 GB/s. This memory bandwidth is pretty good for a smartphone and is on par with many notebooks.

Google does not directly specify the TDP of the Google Tensor G2, so based on the energy consumption and clock frequencies we estimate the TDP to be 10 watts. This puts it in an area in which many modern smartphone processors work.

• Efficient 5-nanometer process
• Strengths in artificial intelligence
• Integrated ARM Mali-G710 MP7 graphics unit
• Solid cache configuration (8 MB L2, 4 MB L3)