Just yesterday, we confirmed the date of the presentation of the Pixel 7, the high-end smartphone that will be the star of the event that Google is preparing for October 6. As usual, we will carefully tell you all the keys of the mentioned terminal, although we can already describe some important details about its SoC, Google Tensor 2.

I’m sure many of our readers remember that with the Pixel 6, Mountain View decided to make a major U-turn on the hardware level. Google ditched Qualcomm’s Snapdragon line and opted for mount a “custom” chip known as Tensor, in quotes because it was actually a Samsung Exynos with a slight layer of customization.

We already told you that the Google Tensor was a high-performance chip that is based on the Samsung Exynos design and that it has 8-core CPU divided into two X1 cores at 2.8 GHz, two Cortex-A76 cores at 2.25 GHz and four low-energy cores at 1.8 GHz. Its GPU is Mali-G78MP20, it has Titan M2 security processor and TPU for AI and deep learning. It is produced in 5nm.

I wanted to share with you the keys to the original Google Tensor, because in general everything indicates that the second generation model will follow in its footsteps. This means that the Mountain View giant will favor brains over brawn, that is, the specialization and value of artificial intelligence and deep learning over raw performance. This makes a lot of sense and is completely understandable, especially when you see how key technologies like AI-powered computational photography are making an impact.

Does this mean that Google Tensor 2 will have average performance? Not at all, but it means it won’t be on the same level than the current top chip, at least in raw performance. This may not be a problem, as today’s upper mid-range chips already “exceed” almost any task and offer excellent performance even in demanding applications.

So what can we expect from Google Tensor 2? Performance improvements at the CPU and GPU level, which puts it on par with current high-end SoCs, and we could also see a new TPU and a new ISP (Image Signal Processor for its acronym in English). It should replicate the 5nm manufacturing node and will most likely give us access to amazing new AI-powered features.