The Tensor G2 is a 5nm Processor, Instead of 4nm That We All Thought it Was

According to the newly surfaced information, the much-hyped Google Tensor G2 is actually built upon the 5nm manufacturing process, and not the 4nm as we were told before.

This means that Tensor G2 found in the Pixel 7 series is not utilizing a 4nm processor. Some confusion surely has arisen from the pre-release benchmarks and the fact that Google decided to not confirm the process used for Tensor G2 during the keynote. Sure, there were claims of 60% speed and 20% efficiency gains over the previous generation but no concrete information was shared.

Much-Hyped Tensor G2 Turns Out to be Built On 5nm Fabrication Process Instead of 4nm

A Google spokesperson actually confirmed to Android Authority that the Tensor G2 is, in fact, built using the 5nm process. Here is the statement.

We purpose-built Google Tensor G2 for real-world use cases. Our final architecture, which includes 5nm, helped us reach that goal while increasing both performance and power efficiency. This approach also allowed us to add new capabilities while taking a step forward on machine learning with our next generation TPU with G2.

As the source notes, this is still not an answer that is fully comprehensive to the question, as Google still has not given the specific manufacturing process used. This does mean that on paper, the 5nm Tensor G2 is not as efficient as the Snapdragon 8 Gen 1 or even Samsung’s Exynos 2200. This might be a disappointment considering how Google hyped the life out of the Tensor G2, and now we are finding out that improvement is not as massive as one would have hoped.

At the same time, it is worth noting that, the real-life performance difference between these devices might not be all that much, in the first place.

Have you ordered a Pixel 7 device? Let us know how you feel about this discovery and whether or not you are okay with it.

The post The Tensor G2 is a 5nm Processor, Instead of 4nm That We All Thought it Was by Furqan Shahid appeared first on Wccftech.