Google Tensor G2 vs Google Tensor: What’s the difference?

Google has today unveiled the Google Tensor G2 chip, which will succeed the existing Google Tensor processor found inside the Pixel 6 and Pixel 6 Pro smartphones.
But what kind of upgrades are you getting with the second-generation chip? We’ve listed all of the major differences right here, so you know what benefits you’re getting by upgrading to the new Pixel 7 and Pixel 7 Pro phones.
Here’s everything you need to know:
The Google Tensor G2 is faster
Let’s start with the most obvious upgrade: the Tensor G2 is the faster of the two. This is thanks to the 4nm manufacturing process, an upgrade from the 5nm node of the original Tensor. The smaller node allows Google to cram more transistors onto the chip, therefore boosting the performance.
Google has introduced new big and medium CPU cores to the chip, with marginally higher frequency speeds. This again, should contribute to the performance boost.
Don’t expect a performance upgrade big enough to challenge rival smartphone chips though! A benchmark leak (reported by NotebookCheck) seemingly indicates that the Tensor G2 still lags behind the speed of the Qualcomm’s Snapdragon 8+ Gen 1 processor.
Improved understanding of speech and language
The Google Tensor G2 has more advanced machine learning than its predecessor, with Google claiming it offers a 60% faster, and 20% more efficient performance in this regard. Google says this helps the Pixel 7 phones to have a better understanding of speech and language.
Thanks to the advancements of the Tensor G2 chip, Pixel 7 phones will be able to transcribe audio messages on Google’s Messages app as soon as you’ve received them. This will be especially useful to those who are hard of hearing.
The Google Assistant voice typing will now be compatible with a greater number of languages too, including French, Italian and Spanish. The original Tensor chip was limited to English, German and Japanese.
And while the ‘Clear Calling’ feature won’t be available on Pixel phones at launch, Google has confirmed that it will eventually help to recognise background noise during a call, and automatically filter out the likes of loud wind and traffic.

Enhanced camera quality
If you’ve ever used a Pixel phone, you’ll know that they’re incredibly good at taking photos. That’s not necessarily due to the camera hardware, but instead because of the incredible advanced machine learning capabilities that the Google Tensor processor allows for.
The Google Tensor G2 is even smarter than its predecessor, and is helping to improve photos even further in several ways.
Firstly, Google claims that low-light photography will be improved since the Tensor G2 offers 2x faster Night Sight processing. The new Pixel phones will also be better at zooming in, provide sharper photos with Face Unblur and will be capable of new features such as Cinematic Blur.
Overall then, the Tensor G2 will make the Pixel 7 phones even better at taking photos and video, which is hugely impressive since the Tensor-powered Pixel 6 series already excelled at camera quality.
The Google Tensor G2 sports a new GPU
Dig into the specs, and you’ll see that the Tensor G2 packs a new GPU called the ARM Mali G710 MP07. For comparison, the original Tensor features a GPU called the ARM Mali G78 MP20.
What are the differences between these two GPUs? Honestly, we’re not sure yet as Google hasn’t really focused on this upgrade during the launch of the Pixel 7 phones.
We’re expecting a minor boost to graphics performance, improving the experience with graphics-intensive workloads such as gaming and video editing. But with Google making minimal fuss about the upgrade, we don’t expect a significant performance gain.