The new processor lands on the market to give life to the new Pixel 7 and Pixel 7 Pro, and possibly to a future Pixel 7a for which we currently have no official information or solid clues. As expected, Google improves all aspects of its chip but focuses its efforts on artificial intelligence. Google knows that it is a key, differentiating factor, and the Pixel 7s put it into practice.
Intelligence above all, even if it is artificial
Google understood last year that its path in the mobile market had to differentiate itself from that of its competition, so it stopped fighting to offer a mobile with the most powerful chip in terms of brute force. Google has always had the artificial intelligence algorithms more advanced so I needed a chip designed ex profeso for them. And that’s how the original Google Tensor was born.
With the arrival of Google Tensor 2, manufacturing technology is reduced. We leave 5 nanometers behind to move to 4 nanometer lithography and with this we gain energy efficiency but also power. If everything is smaller, the density of transistors can be higher. But it has not been the only change. Google has also improved the CPU by introducing more efficient and powerful cores.
If in the original Tensor we had a 2+2+4 configuration with Cortex X1, Cortex A76 and Cortex A55, now we have Cortex X1, Cortex A78 and Cortex A55. The A78s also run at 2.35GHz instead of 2.25GHz, and the Cortex X1s go up to 2.85GHz. Perhaps somewhat continuous changes, but important when it comes to offering more power. The Mali G78 has been superseded by the Mali G710, by the way. More graphics power.
However, the crown jewel of the Google Tensor G2 lies in its ability to handle artificial intelligence, code machine learning, and there Google has given everything. According to the company itself, this AI code now renders 60% fasterBe careful with this data that is not insignificant, thanks to the arrival of its new TPU, or Tensor Processing Unit. The Pixel NPU, to understand us.
We do not forget the energy consumption since Google tells us that the Tensor G2 is 20% more efficient than the original Tensor, which should result in greater autonomy. In addition, the arrival of the processor makes voice recognition faster and more powerful. Something that applies, for example, to the automatic transcription of messages through Assistant. This system is also used for ‘Clear Calling’ which allows calls to be improved by erasing background noise, such as wind or traffic noise.
And of course, the artificial intelligence of the Tensor G2 is also applied to the photography of the Pixel 7 and Pixel 7 Pro. Night photographic processing is improved, the Face Unblur approach is faster and more powerful and Cinematic Blur comes to video. And yes, artificial intelligence is also used for the new Super Zoom, and for the focus and processing of macro images. All this is what comes with a Tensor G2 that, as usual, It will be limited solely and exclusively to the seventh generation of Google Pixel mobiles.