Google Tensor G4 vs Google Tensor G5

The Google Tensor G4, based on the G4 architecture, is manufactured using a 4 nm fabrication process.

Its successor, the Google Tensor G5, utilizes the G5 Laguna architecture and the more advanced 3 nm fabrication process.

Both processors use ARM core architectures. Details regarding core/thread configurations are not available. In terms of computing power, the Google Tensor G5 shows a clear superiority. In Geekbench 6, it outperforms the Google Tensor G4 in single-core performance by over 22 percent and in multi-core performance by more than 40 percent.

This indicates a significantly increased computational capability. For gaming scenarios, the Google Tensor G5, with its IMG DXT-48-1536 graphics solution, offers advantages over the ARM Immortalis-G715 MP10 of the Google Tensor G4. The smaller 3 nm fabrication process and the newer graphics architecture of the G5 promise better graphics performance.

The memory bandwidth is 51 GB/s for both. Specific information on Thermal Design Power (TDP) is missing. However, the switch to the 3 nm fabrication process in the Google Tensor G5 implies improved energy efficiency, which enables higher performance at the same power consumption. As integrated SoC components, the Google Tensor G4 and Google Tensor G5 are not available as separate desktop processors on the open market.

A direct evaluation of price-performance ratio or market prices is therefore not possible. In summary, the Google Tensor G5 offers superior performance in all mentioned areas, especially in computing and graphics performance. The Google Tensor G4 provides solid basic performance.

For the highest demands, the Google Tensor G5 is the preferred choice; for less extreme requirements, the Google Tensor G4 could be a suitable solution.