The chip, which equips the latest Pixel 7 and Pixel 7 Prodoes not shine particularly in the benchmarks, but this is not the objective of Google.
In 2021, Google launched its first mobile processor, the Google Tensor which powered the Pixel 6 and Pixel 6 Pro.
A disappointing Google Tensor G2, but not for the manufacturer
The American company had therefore chosen to free itself from a Qualcomm by designing its own chips, thought out and optimized for its own needs. Rebelote this year with the Google Tensor G2, a slightly more powerful chip than the previous one that equips the latest smartphones from the manufacturer.
The Pixel 7 and Pixel 7 Pro have obviously undergone the test of benchmarks, and the results obtained are a bit disappointing. The Google Tensor G2 is outpaced both by Apple’s A16 Bionic processors, but also by several chips used by many Android smartphones, such as Qualcomm’s Snapdragon 8+ Gen 1.
Asked about this during a podcast, Monika Gupta, product manager at Google Silicon (the division responsible for designing processors) kicked in and indicated that her teams were ” perfectly comfortable with the idea that the Google Tensors would never rise to the top of the benchmark rankings.
Google relies on AI, more than raw power
For the manager, raw performance is not the main reason that guides Google in its design choices. The brand prefers to focus on its work in artificial intelligence:
” I think classic benchmarks served a purpose at one time, but I think the industry has evolved since then. Look at what Google is trying to do by pushing AI innovations into a smartphone, because we think that’s the approach that will deliver useful experiences. [….] Benchmarks were created in a time when AI and phones didn’t even exist. »
Google values more what it calls “Ambient Technology”, or devices and systems that work together to provide the right information at the right time and improve the lives of users.
” I’d say we’re building on that vision of ambient computing and figuring out how to do super complex and nuanced things on-chip in a power-efficient way. concludes Monika Gupta.
The division of Google dedicated to processors is already working on the future of Google Tensor, with a development plan that covers the next five years.
Source : Wccftech
12