Google’s Tensor chip has, without doubt, made the headlines as the new Pixel 6 series has now been announced. The chip has indeed long been marketed as being big on machine learning as well as the process of supplying all the power and juice that happens to be required for stuff like Google’s improved computational photography and transcription abilities.
Pixel fans have since been arguing that while the Tensor chip does perhaps lack a bit in performance, it compensates this thanks to its ML prowess. For those of you who are uninitiated, the Pixel 6 is indeed outperformed by both the Snapdragon 888 as well as the Exynos 2100 phones and also Apple’s A15 and even the A14 Bionic chips.
However, in a new test which was happened to have been performed by some researchers, it was actually found that the A15 Bionic even managed to crush the Google tensor Chip in its specified area of expertise : machine learning. Indeed the test which was carried out was done so as to specifically measure machine learning performance.
Also Read: Hyundai to soon develop chips itself
The Google Tensor managed a score of only 307, while Apple’s A15 managed a very impressive score of 939. While the 307 score achieved by the Tensor chip isn’t bad by any stretch of the imagination, it just goes on to show that Google’s chip is no match to Apple’s even though it’s in an area where the Tensor poses strength.
Also, it’s important to consider the fact that this chip is Google’s very first self-developed chip so one would have to imagine that it’s only deserved that we all cut Google some slack. Moreover, this chip actually managed to beat both the Snapdragon 888 and the Exynos 2100 as well as the Kirin 9000 in the GPU benchmark tests – all this also in accordance with what the report suggests!
