Trusted Reviews is supported by its audience. If you purchase through links on our site, we may earn a commission. Learn more.

What is Google Tensor G2?

Google has unveiled the Tensor G2 chip for its Pixel 7 smartphone series. But what is this chip, and how does it compare to the preceding Tensor processor.

Here’s everything you need to know about the Google Tensor G2 chipset.

What is Google Tensor G2?

The Google Tensor G2 is the second-generation smartphone processor by Google that is designed for the Pixel range.

The Pixel 6 and Pixel 6 Pro were the first phones to adopt the original Tensor chip, and now Google has confirmed the Pixel 7 and Pixel 7 Pro will be the first phones to be powered by the new Tensor G2 chip.

The G2 is built on new architecture, which seemingly uses Samsung’s 4nm manufacturing process. This is an improvement on the 5nm node of the original Tensor chip, as it should allow Google to squeeze more transistors onto the processor to enhance the performance.

Judging by the official specs below, Google has also introduced new big and and medium CPU cores, providing minor boosts to the frequency speeds. We don’t know what kind of performance gains to expect just yet, but it will inevitably be faster than the original.

Tensor G2Tensor
Manufacturing process4nm5nm
Big CPU cores2x 2.85GHz Cortex-X12x 2.8GHz Cortex-X1
Medium CPU cores2x 2.35GHz Cortex-A782x 2.25GHz Cortex-A76
Small CPU Cores4x 1.8GHz Cortex-A554x 1.8GHz Cortex-A55
GPUARM Mali G710 MP07ARM Mali G78 MP20

Google has also introduced a new GPU to the Pixel processor, which should see a boost to graphics-intensive workloads such as gaming.

But just like with the original Tensor processor, it’s the new features via the improved advanced machine learning (rather than the raw performance gains) that makes the G2 an exciting upgrade. Google claims the new G2 chip offers a 60% faster and 20% more efficient machine-learning performance compared to the original Tensor chip.

Google has highlighted the improved speech recognition as a major upgrade, with the new Pixel 7 phones able to automatically transcribe audio messages on Google’s Messages app as soon as you receive them. Google Assistant voice typing is now available in French, Italian and Spanish, as well as English, German, and Japanese. And you can even search for emoji via your voice.

And while this feature won’t be available at launch, the Google’s new chip will allow the Pixel phones to use ‘Clear Calling’, which uses Google’s machine learning to filter out distracting background noise such as loud wind and traffic.

The smarts of the Google Tensor G2 helps to improve your camera too. Upgraded camera features include next-gen Super Zoom, 2x faster Night Sight processing, the new Cinematic Blur and sharper photos with Face Unblur.

With all this in mind, it’s important to remember that the main appeal of the Google Tensor G2 chip is its use of machine learning to improve the experience of day-to-day tasks such as taking a photo or messaging a friend. The performance boost should only really be considered a bonus rather than the main focus.

Why trust our journalism?

Founded in 2003, Trusted Reviews exists to give our readers thorough, unbiased and independent advice on what to buy.

Today, we have millions of users a month from around the world, and assess more than 1,000 products a year.

author icon

Editorial independence

Editorial independence means being able to give an unbiased verdict about a product or company, with the avoidance of conflicts of interest. To ensure this is possible, every member of the editorial staff follows a clear code of conduct.

author icon

Professional conduct

We also expect our journalists to follow clear ethical standards in their work. Our staff members must strive for honesty and accuracy in everything they do. We follow the IPSO Editors’ code of practice to underpin these standards.

Trusted Reviews Logo

Sign up to our newsletter

Get the best of Trusted Reviews delivered right to your inbox.

This is a test error message with some extra words