How Pixel 7’s Tensor G2 Chip Could Compare to A16 Bionic

BY Sriansh

Published 19 Sep 2022

Google Tensor G2 SoC

Google is all set to introduce its Pixel 7 series next month. The company has announced that it will unveil its Pixel 7 and Pixel 7 Pro at an event on October 6. Along with this, the search engine giant has also revealed that the Pixel 7 series will be powered by the second-generation Tensor G2 chipset. Now a leaked benchmark of the Tensor G2 has surfaced, giving us an idea of how the new Google chipset would compare with Apple’s A16 Bionic SoC.

Popular Android developer Kuba Wojciechowski says Google Tensor G2 will feature a similar CPU as the first-generation chipset. This means that the chipset will come with two Cortex-X1 cores, two Cortex-A76 cores, and four Cortex-A55 cores. The cores, however, are clocked at a slightly higher speed this time compared to before. 

According to the Geekbench listing, which 9to5Google believes to be ‘legitimate,’ the Pixel 7 Pro achieves a single-core score of 1068 and a multi-core score of 3149. While the Tensor G2 brings a 10 percent improvement over the previous generation, it is still lightyears behind Apple’s A16 Bionic, which achieved a score of 1891 in the single-core test and 5469 points in the multi-core test.

While the CPU improvements may have disappointed, it seems that Google Tensor G2 will bring advancements in other areas. According to Kuba, Google has worked on heat reduction and thermal efficiency of the chipset, and now the Tensor chipset can run at higher clock speeds for longer periods of time. 

In addition, the Tensor G2 will bring significant improvements in the graphics department, thanks to the new Mali-G710 GPU. There will be a 20% improvement in performance, a 20% improvement in power efficiency, and a 35% improvement in machine learning performance — resulting in a significantly improved Google Camera.

While the Tensor G2’s GPU and CPU performance may appear weaker compared to Apple’s A16 Bionic, Google’s chipset has always given Apple tough competition in the machine learning department. Google was the first smartphone maker to implement offline dictation on its smartphone, thanks to the chipset’s TPU (Tensor Processing Unit). A similar feature was then implemented by Apple with iOS 16

Apple introduced A16 Bionic along with iPhone 14 and 14 Pro Max earlier this month. Even though the chipset features only six cores, containing two high-performance cores and four high-efficiency cores, it still performs better compared to its competitor chipsets. Apple, during the iPhone 14 launch keynote, claimed that the CPU of the A16 Bionic is up to 40 percent faster than rivals and up to 50 percent better in terms of graphics performance.

While it is too early to judge Tensor G2’s performance, it seems that it will take some time for Google to catch up to Apple in this department. What are your thoughts on the leaked Tensor G2 SoC benchmarks? Do you think benchmark scores are an accurate reflection of a chipset’s performance capabilities? Let us know in the comments section below!

Source: Twitter, Geekbench | Via: 9to5Google